01_Solver.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "6b0fd06a-f3cb-4732-bc7d-1c7ad2d39c0f",
   "metadata": {},
   "source": [
    "# Comparison Loss Function and Validation Score"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5bce52fe-8793-4ee5-8b75-1932fd4aa55d",
   "metadata": {},
   "source": [
    "1) Import der einzelnen Libraries. Sortieren der Daten nach der Zielkategorie für den ersten Überblick.\n",
    "2) Aufteilen der Trainings- und Testdaten. Erstellen von notwendigen Listen\n",
    "3) Erstellen von Funktionen \n",
    "4) initialisieren der einzelnen MLPClassifier Netze \n",
    "5) Ergebnisse: Textuelle Daten, geplottete Vergleiche, prozentuale Heatmaps und Reports"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e3990afb-733a-40a5-957e-451abceb2d5e",
   "metadata": {},
   "source": [
    "## imports and data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "e9ce620e-67af-4fcb-8736-ead8893c3523",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['data', 'target', 'frame', 'target_names', 'feature_names', 'DESCR'])"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.datasets import fetch_covtype\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.neural_network import MLPClassifier\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import classification_report,confusion_matrix, balanced_accuracy_score, roc_auc_score, top_k_accuracy_score, cohen_kappa_score, matthews_corrcoef, hinge_loss\n",
    "from sklearn.metrics import plot_confusion_matrix, multilabel_confusion_matrix\n",
    "from sklearn.metrics import RocCurveDisplay, accuracy_score, recall_score, precision_score, precision_recall_fscore_support\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.figure as fig\n",
    "from matplotlib.colors import ListedColormap\n",
    "from mpl_toolkits import mplot3d\n",
    "import seaborn as sns\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "covertype = fetch_covtype()\n",
    "forest = fetch_covtype(as_frame=True)\n",
    "covertype.keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "a574498c-78cc-4eb5-9626-ea65dab5307e",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0         5\n",
       "1         5\n",
       "2         2\n",
       "3         2\n",
       "4         5\n",
       "         ..\n",
       "581007    3\n",
       "581008    3\n",
       "581009    3\n",
       "581010    3\n",
       "581011    3\n",
       "Name: Cover_Type, Length: 581012, dtype: int32"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "forest.target"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "3a19f2f3-1f0e-43f7-88fd-f3a9b3b7bdf9",
   "metadata": {},
   "outputs": [],
   "source": [
    "forestlist = forest.frame\n",
    "forestlistFiltered = forestlist.filter(items=['Elevation', 'Aspect', 'Slope', 'Horizontal_Distance_To_Hydrology', 'Vertical_Distance_To_Hydrology', 'Horizontal_Distance_To_Roadways', 'Hillshade_9am', 'Hillshade_Noon', 'Hillshade_3pm', 'Horizontal_Distance_To_Fire_Points', 'Wilderness_Area_0', 'Wilderness_Area_1', 'Wilderness_Area_2', 'Wilderness_Area_3', 'Cover_Type'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "4ec34c71-e117-44eb-a8bb-1d7acdd756f5",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Elevation</th>\n",
       "      <th>Aspect</th>\n",
       "      <th>Slope</th>\n",
       "      <th>Horizontal_Distance_To_Hydrology</th>\n",
       "      <th>Vertical_Distance_To_Hydrology</th>\n",
       "      <th>Horizontal_Distance_To_Roadways</th>\n",
       "      <th>Hillshade_9am</th>\n",
       "      <th>Hillshade_Noon</th>\n",
       "      <th>Hillshade_3pm</th>\n",
       "      <th>Horizontal_Distance_To_Fire_Points</th>\n",
       "      <th>...</th>\n",
       "      <th>Soil_Type_31</th>\n",
       "      <th>Soil_Type_32</th>\n",
       "      <th>Soil_Type_33</th>\n",
       "      <th>Soil_Type_34</th>\n",
       "      <th>Soil_Type_35</th>\n",
       "      <th>Soil_Type_36</th>\n",
       "      <th>Soil_Type_37</th>\n",
       "      <th>Soil_Type_38</th>\n",
       "      <th>Soil_Type_39</th>\n",
       "      <th>Cover_Type</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>290505</th>\n",
       "      <td>2928.0</td>\n",
       "      <td>58.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>150.0</td>\n",
       "      <td>-7.0</td>\n",
       "      <td>1836.0</td>\n",
       "      <td>226.0</td>\n",
       "      <td>225.0</td>\n",
       "      <td>133.0</td>\n",
       "      <td>190.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>532739</th>\n",
       "      <td>3015.0</td>\n",
       "      <td>343.0</td>\n",
       "      <td>18.0</td>\n",
       "      <td>376.0</td>\n",
       "      <td>50.0</td>\n",
       "      <td>3378.0</td>\n",
       "      <td>181.0</td>\n",
       "      <td>211.0</td>\n",
       "      <td>166.0</td>\n",
       "      <td>1982.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>532740</th>\n",
       "      <td>3018.0</td>\n",
       "      <td>341.0</td>\n",
       "      <td>18.0</td>\n",
       "      <td>350.0</td>\n",
       "      <td>53.0</td>\n",
       "      <td>3398.0</td>\n",
       "      <td>180.0</td>\n",
       "      <td>211.0</td>\n",
       "      <td>168.0</td>\n",
       "      <td>2005.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>532741</th>\n",
       "      <td>3021.0</td>\n",
       "      <td>345.0</td>\n",
       "      <td>18.0</td>\n",
       "      <td>324.0</td>\n",
       "      <td>45.0</td>\n",
       "      <td>3419.0</td>\n",
       "      <td>182.0</td>\n",
       "      <td>209.0</td>\n",
       "      <td>164.0</td>\n",
       "      <td>2029.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>532742</th>\n",
       "      <td>3023.0</td>\n",
       "      <td>349.0</td>\n",
       "      <td>19.0</td>\n",
       "      <td>300.0</td>\n",
       "      <td>47.0</td>\n",
       "      <td>3439.0</td>\n",
       "      <td>182.0</td>\n",
       "      <td>206.0</td>\n",
       "      <td>160.0</td>\n",
       "      <td>2052.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>399826</th>\n",
       "      <td>3334.0</td>\n",
       "      <td>356.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>558.0</td>\n",
       "      <td>100.0</td>\n",
       "      <td>765.0</td>\n",
       "      <td>204.0</td>\n",
       "      <td>223.0</td>\n",
       "      <td>156.0</td>\n",
       "      <td>1642.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>399825</th>\n",
       "      <td>3333.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>601.0</td>\n",
       "      <td>99.0</td>\n",
       "      <td>806.0</td>\n",
       "      <td>212.0</td>\n",
       "      <td>226.0</td>\n",
       "      <td>152.0</td>\n",
       "      <td>1583.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>399824</th>\n",
       "      <td>3334.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>601.0</td>\n",
       "      <td>39.0</td>\n",
       "      <td>845.0</td>\n",
       "      <td>214.0</td>\n",
       "      <td>231.0</td>\n",
       "      <td>155.0</td>\n",
       "      <td>1524.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>399812</th>\n",
       "      <td>3331.0</td>\n",
       "      <td>246.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>660.0</td>\n",
       "      <td>117.0</td>\n",
       "      <td>859.0</td>\n",
       "      <td>213.0</td>\n",
       "      <td>241.0</td>\n",
       "      <td>166.0</td>\n",
       "      <td>1113.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>293238</th>\n",
       "      <td>3284.0</td>\n",
       "      <td>307.0</td>\n",
       "      <td>18.0</td>\n",
       "      <td>60.0</td>\n",
       "      <td>22.0</td>\n",
       "      <td>3922.0</td>\n",
       "      <td>167.0</td>\n",
       "      <td>226.0</td>\n",
       "      <td>198.0</td>\n",
       "      <td>3999.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>581012 rows × 55 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        Elevation  Aspect  Slope  Horizontal_Distance_To_Hydrology  \\\n",
       "290505     2928.0    58.0    7.0                             150.0   \n",
       "532739     3015.0   343.0   18.0                             376.0   \n",
       "532740     3018.0   341.0   18.0                             350.0   \n",
       "532741     3021.0   345.0   18.0                             324.0   \n",
       "532742     3023.0   349.0   19.0                             300.0   \n",
       "...           ...     ...    ...                               ...   \n",
       "399826     3334.0   356.0   10.0                             558.0   \n",
       "399825     3333.0     6.0    7.0                             601.0   \n",
       "399824     3334.0     3.0    5.0                             601.0   \n",
       "399812     3331.0   246.0    3.0                             660.0   \n",
       "293238     3284.0   307.0   18.0                              60.0   \n",
       "\n",
       "        Vertical_Distance_To_Hydrology  Horizontal_Distance_To_Roadways  \\\n",
       "290505                            -7.0                           1836.0   \n",
       "532739                            50.0                           3378.0   \n",
       "532740                            53.0                           3398.0   \n",
       "532741                            45.0                           3419.0   \n",
       "532742                            47.0                           3439.0   \n",
       "...                                ...                              ...   \n",
       "399826                           100.0                            765.0   \n",
       "399825                            99.0                            806.0   \n",
       "399824                            39.0                            845.0   \n",
       "399812                           117.0                            859.0   \n",
       "293238                            22.0                           3922.0   \n",
       "\n",
       "        Hillshade_9am  Hillshade_Noon  Hillshade_3pm  \\\n",
       "290505          226.0           225.0          133.0   \n",
       "532739          181.0           211.0          166.0   \n",
       "532740          180.0           211.0          168.0   \n",
       "532741          182.0           209.0          164.0   \n",
       "532742          182.0           206.0          160.0   \n",
       "...               ...             ...            ...   \n",
       "399826          204.0           223.0          156.0   \n",
       "399825          212.0           226.0          152.0   \n",
       "399824          214.0           231.0          155.0   \n",
       "399812          213.0           241.0          166.0   \n",
       "293238          167.0           226.0          198.0   \n",
       "\n",
       "        Horizontal_Distance_To_Fire_Points  ...  Soil_Type_31  Soil_Type_32  \\\n",
       "290505                               190.0  ...           1.0           0.0   \n",
       "532739                              1982.0  ...           0.0           0.0   \n",
       "532740                              2005.0  ...           0.0           0.0   \n",
       "532741                              2029.0  ...           0.0           0.0   \n",
       "532742                              2052.0  ...           0.0           0.0   \n",
       "...                                    ...  ...           ...           ...   \n",
       "399826                              1642.0  ...           0.0           0.0   \n",
       "399825                              1583.0  ...           1.0           0.0   \n",
       "399824                              1524.0  ...           1.0           0.0   \n",
       "399812                              1113.0  ...           1.0           0.0   \n",
       "293238                              3999.0  ...           0.0           0.0   \n",
       "\n",
       "        Soil_Type_33  Soil_Type_34  Soil_Type_35  Soil_Type_36  Soil_Type_37  \\\n",
       "290505           0.0           0.0           0.0           0.0           0.0   \n",
       "532739           0.0           0.0           0.0           0.0           0.0   \n",
       "532740           0.0           0.0           0.0           0.0           0.0   \n",
       "532741           0.0           0.0           0.0           0.0           0.0   \n",
       "532742           0.0           0.0           0.0           0.0           0.0   \n",
       "...              ...           ...           ...           ...           ...   \n",
       "399826           0.0           0.0           0.0           0.0           0.0   \n",
       "399825           0.0           0.0           0.0           0.0           0.0   \n",
       "399824           0.0           0.0           0.0           0.0           0.0   \n",
       "399812           0.0           0.0           0.0           0.0           0.0   \n",
       "293238           0.0           0.0           0.0           0.0           1.0   \n",
       "\n",
       "        Soil_Type_38  Soil_Type_39  Cover_Type  \n",
       "290505           0.0           0.0           1  \n",
       "532739           0.0           0.0           1  \n",
       "532740           0.0           0.0           1  \n",
       "532741           0.0           0.0           1  \n",
       "532742           0.0           0.0           1  \n",
       "...              ...           ...         ...  \n",
       "399826           0.0           0.0           7  \n",
       "399825           0.0           0.0           7  \n",
       "399824           0.0           0.0           7  \n",
       "399812           0.0           0.0           7  \n",
       "293238           0.0           0.0           7  \n",
       "\n",
       "[581012 rows x 55 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "forestlist = forest.frame\n",
    "forestlistSorted = forestlist.sort_values(by='Cover_Type')\n",
    "forestlistSorted"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "47013291-f624-41ea-b385-6a707c06a97f",
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'plotData' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m/tmp/ipykernel_2906648/1355288166.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mplotData\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m: name 'plotData' is not defined"
     ]
    }
   ],
   "source": [
    "plotData()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ad685848-0852-40ec-9224-6479edcfbadf",
   "metadata": {
    "tags": []
   },
   "source": [
    "## training and test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "d7374455-6ab6-482b-914d-7d25c028307e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(581012, 54)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "covertype['data'].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "aa2226cd-ac5d-44b9-81d3-86e117c60d97",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_c = covertype['data']\n",
    "y_c = covertype['target']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "e95d3764-9b2c-4c58-a730-8b7167b298a5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(435759, 54)\n",
      "(145253, 54)\n"
     ]
    }
   ],
   "source": [
    "X_train_c, X_test_c, y_train_c, y_test_c = train_test_split(X_c, y_c)\n",
    "print(X_train_c.shape) \n",
    "print(X_test_c.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "1e55d24c-b370-4bec-95d4-52c00074d70c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "StandardScaler()"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scaler_c = StandardScaler()\n",
    "# Fit only to the training data\n",
    "scaler_c.fit(X_train_c)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "2892e8de-0ce9-4e7a-aab7-6b188f2a8c0d",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train_c = scaler_c.transform(X_train_c)\n",
    "X_test_c = scaler_c.transform(X_test_c)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "fbdcb1ab-db1f-4585-9d11-b7c56775370d",
   "metadata": {},
   "outputs": [],
   "source": [
    "classificationList = []\n",
    "curve = []\n",
    "validation = []\n",
    "iteration = []\n",
    "\n",
    "#validationList = []\n",
    "#curve_test = []\n",
    "#validation_test = []\n",
    "#iteration_test = []\n",
    "\n",
    "accuracyList = []\n",
    "recallList = []\n",
    "precisionList = []"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1ad8a3ff-1ec6-4b47-b833-497c31f505d7",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "## functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "00451c60-04e7-4cb1-85f2-cc748c4d0ba4",
   "metadata": {},
   "outputs": [],
   "source": [
    "def plotData():\n",
    "    fig, axes = plt.subplots(3,2)\n",
    "    sns.histplot(data=forestlistSorted, x=\"Elevation\", color=\"skyblue\", label=\"Elevation\", kde=True, ax=axes[0,0])\n",
    "    sns.histplot(data=forestlistSorted, x=\"Aspect\", color=\"red\", label=\"Aspect\", kde=True, ax=axes[0,0])\n",
    "    sns.histplot(data=forestlistSorted, x=\"Slope\", color=\"orange\", label=\"Slope\", kde=True, ax=axes[0,0])\n",
    "    plt.legend() \n",
    "\n",
    "    sns.histplot(data=forestlistSorted, x=\"Horizontal_Distance_To_Hydrology\", color=\"skyblue\", label=\"Horizontal_Distance_To_Hydrology\", kde=True, ax=axes[0,1])\n",
    "    sns.histplot(data=forestlistSorted, x=\"Vertical_Distance_To_Hydrology\", color=\"red\", label=\"Vertical_Distance_To_Hydrology\", kde=True, ax=axes[0,1])\n",
    "    sns.histplot(data=forestlistSorted, x=\"Horizontal_Distance_To_Roadways\", color=\"orange\", label=\"Horizontal_Distance_To_Roadways\", kde=True, ax=axes[0,1])\n",
    "    sns.histplot(data=forestlistSorted, x=\"Horizontal_Distance_To_Fire_Points\", color=\"green\", label=\"Horizontal_Distance_To_Fire_Points\", kde=True, ax=axes[0,1])\n",
    "    plt.legend() \n",
    "\n",
    "    sns.histplot(data=forestlistSorted, x=\"Hillshade_9am\", color=\"skyblue\", label=\"Hillshade_9am\", kde=True, ax=axes[1,0])\n",
    "    sns.histplot(data=forestlistSorted, x=\"Hillshade_Noon\", color=\"red\", label=\"Hillshade_Noon\", kde=True, ax=axes[1,0])\n",
    "    sns.histplot(data=forestlistSorted, x=\"Hillshade_3pm\", color=\"orange\", label=\"Hillshade_3pm\", kde=True, ax=axes[1,0])\n",
    "    plt.legend() \n",
    "\n",
    "    sns.histplot(data=forestlistSorted, x=\"Soil_Type_31\", color=\"skyblue\", label=\"Soil_Type_31\", kde=True, ax=axes[1,1])\n",
    "    sns.histplot(data=forestlistSorted, x=\"Soil_Type_32\", color=\"red\", label=\"Soil_Type_32\", kde=True, ax=axes[1,1])\n",
    "    sns.histplot(data=forestlistSorted, x=\"Soil_Type_33\", color=\"orange\", label=\"Soil_Type_33\", kde=True, ax=axes[1,1])\n",
    "    sns.histplot(data=forestlistSorted, x=\"Soil_Type_34\", color=\"cyan\", label=\"Soil_Type_34\", kde=True, ax=axes[1,1])\n",
    "    sns.histplot(data=forestlistSorted, x=\"Soil_Type_35\", color=\"darkgreen\", label=\"Soil_Type_35\", kde=True, ax=axes[1,1])\n",
    "    sns.histplot(data=forestlistSorted, x=\"Soil_Type_36\", color=\"gold\", label=\"Soil_Type_36\", kde=True, ax=axes[1,1])\n",
    "    sns.histplot(data=forestlistSorted, x=\"Soil_Type_37\", color=\"purple\", label=\"Soil_Type_37\", kde=True, ax=axes[1,1])\n",
    "    sns.histplot(data=forestlistSorted, x=\"Soil_Type_38\", color=\"silver\", label=\"Soil_Type_38\", kde=True, ax=axes[1,1])\n",
    "    sns.histplot(data=forestlistSorted, x=\"Soil_Type_39\", color=\"navy\", label=\"Soil_Type_39\", kde=True, ax=axes[1,1])\n",
    "    plt.legend() \n",
    "\n",
    "    sns.histplot(data=forestlistSorted, x=\"Cover_Type\", color=\"orange\", label=\"Cover_Type\",  ax=axes[2,1], discrete=True, kde=False)\n",
    "    plt.legend() \n",
    "    fig.tight_layout()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "207ed85a-e659-4bb9-b533-fc3dc2a39ea1",
   "metadata": {},
   "outputs": [],
   "source": [
    "def addToList(mlp, liste):\n",
    "    liste.append(mlp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "2b80fe74-be82-444d-ac56-b6a8f4797089",
   "metadata": {},
   "outputs": [],
   "source": [
    "def removeFromList(mlp, liste):\n",
    "    liste.remove(mlp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "8384363b-3a23-4ea7-9ddf-e8e2fb0fcd49",
   "metadata": {},
   "outputs": [],
   "source": [
    "def comparison(List):\n",
    "    for mlp in List:\n",
    "        print(mlp)\n",
    "        print('\\nparamters: ' + str(mlp.get_params(deep=True)))\n",
    "        print('\\nlast loss: ' + str(mlp.loss_))\n",
    "        print('\\nnumber of features seen during fit: ' + str(mlp.n_features_in_))\n",
    "        print('\\nnumber of layers: ' + str(mlp.n_layers_))\n",
    "        print('\\nnumber of outputs: ' + str(mlp.n_outputs_))\n",
    "        print('\\nname of output activation function: ' + str(mlp.out_activation_) )\n",
    "        print('\\nnumber of iterations: ' +str(mlp.n_iter_) )\n",
    "        print('\\nnumber training samples seen by the solver during fitting: ' +str(mlp.t_) )\n",
    "        print('****************************************************************************************')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "76d31667-65c3-4d58-89a5-0c8e333620a2",
   "metadata": {},
   "outputs": [],
   "source": [
    "def curveValidateIteration(List, c, v, i):\n",
    "    for mlp in List:\n",
    "        c.append(mlp.loss_curve_)\n",
    "        v.append(mlp.validation_scores_)\n",
    "        i.append(np.arange(1, mlp.n_iter_+1))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ecba6578-cde2-4e01-970c-209a788ea8ed",
   "metadata": {},
   "source": [
    "__________________________________________________________________________________________________________________________________"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "4e8f31d2-0b07-4411-8bc6-1c2c68b140fa",
   "metadata": {},
   "outputs": [],
   "source": [
    "cmap = ListedColormap([\"coral\", \"azure\", \"indigo\", \"darkgreen\", \"lightblue\", \"crimson\", \"gold\", \"silver\", \"teal\"])\n",
    "def plotCurvePerInteration(List, curvy, iterate):\n",
    "    print('default: hidden_layer = (100,), activation = relu, solver = adam, alpha = 0.0001, learning_rate = constant, learning_rate_init = 0.001, max_iter = 200, shuffle = true, early_stopping & warm_start = FALSE')\n",
    "    print('Alle Infos, die davon abweichen, werden in der Legende gelistet')\n",
    "    print('****************************************************************************************')\n",
    "    for i,c,cl in zip(iterate, curvy, List):\n",
    "        plt.scatter(i,c, marker = '.', label = str(cl))\n",
    "    plt.title('Losscurve per iteration')\n",
    "    plt.xlabel(\"iteration\")\n",
    "    plt.ylabel(\"loss curve\")\n",
    "    plt.legend(bbox_to_anchor=(1.05, 1.0), loc='upper left')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "a3ef3f87-42fb-46b9-96f7-7b363ec85fc4",
   "metadata": {},
   "outputs": [],
   "source": [
    "def plotValidationPerInteration(List, validate, iterate):\n",
    "    for i,v, cl in zip(iterate, validate, List):\n",
    "        plt.scatter(i,v, marker = '.', label = str(cl))\n",
    "    plt.title('Validation per iteration')\n",
    "    plt.xlabel(\"iteration\")\n",
    "    plt.ylabel(\"validation\")\n",
    "    plt.legend(bbox_to_anchor=(1.05, 1.0), loc='upper left')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "25ada7f9-9a81-4a58-80ae-1bea09db70f0",
   "metadata": {},
   "outputs": [],
   "source": [
    "def plotValidationPercurve(List, validate, curvy):\n",
    "    for c,v,cl in zip(validate, curvy, List):\n",
    "        plt.scatter(c,v, marker = '.', label = str(cl))\n",
    "    plt.title('Validation per curve')\n",
    "    plt.xlabel(\"validation\")\n",
    "    plt.ylabel(\"curve\")\n",
    "    plt.legend(bbox_to_anchor=(1.05, 1.0), loc='upper left')\n",
    "    plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9fcda0d6-d713-4584-ac38-41b1814e1eec",
   "metadata": {},
   "source": [
    "__________________________________________________________________________________________________________________________________"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "ad8cf6d8-5f4f-477f-9b18-8acb297611fa",
   "metadata": {},
   "outputs": [],
   "source": [
    "def heatmaping(List):\n",
    "    for mlp in List:\n",
    "        print(mlp)\n",
    "        predictions = mlp.predict(X_test_c)   \n",
    "        con_ma = confusion_matrix(y_test_c,predictions)\n",
    "        con_ma_perc = np.empty([7, 7], dtype=float)\n",
    "\n",
    "        for i in range(0, 7):\n",
    "            number = 0.0\n",
    "            for j in range(0, 7):\n",
    "                number += con_ma[i][j]\n",
    "            for j in range(0, 7):\n",
    "                con_ma_perc[i][j] = 100/number*con_ma[i][j]\n",
    "\n",
    "        tree = ['Spruce', 'Lodgepole Pine', 'PoderosaPine', 'Cottonwood/Willow', 'Aspen', 'Douglas-fir', 'Krummholz']\n",
    "\n",
    "        # ======== Plot confuse matrix prozentual ========\n",
    "\n",
    "        ay = sns.heatmap(con_ma_perc, annot=True, cmap='coolwarm', yticklabels=tree, xticklabels=tree, fmt='.2f', vmin=0, vmax=100)\n",
    "        ay.xaxis.tick_top()\n",
    "        ay.xaxis.set_label_position('top')\n",
    "        plt.xticks(rotation=90)\n",
    "        ay.set(title=\"Confusion Matrix - Recall\",\n",
    "              xlabel=\"- Predicted Tree -\",\n",
    "              ylabel=\"- Actual Tree -\")\n",
    "        plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "633860b1-fd10-4b7a-b001-8bacfb1ce97c",
   "metadata": {},
   "outputs": [],
   "source": [
    "def reporting(List):\n",
    "    for mlp in List:\n",
    "        print(mlp)\n",
    "        predictions = mlp.predict(X_test_c)\n",
    "        print(classification_report(y_test_c,predictions))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "1e4d2697-5eb8-4deb-a58a-a9cde3fc2599",
   "metadata": {},
   "outputs": [],
   "source": [
    "def accuracyPlotting(List):\n",
    "        for mlp in List:\n",
    "            print(mlp)\n",
    "            predictions = mlp.predict(X_test_c)\n",
    "            accuracy = accuracy_score(y_test_c, predictions)\n",
    "            recall = recall_score(y_test_c, predictions, average = 'weighted')\n",
    "            precision = precision_score(y_test_c, predictions, average = 'weighted')\n",
    "            \n",
    "            addToList(accuracy, accuracyList)\n",
    "            addToList(recall, recallList)\n",
    "            addToList(precision, precisionList)\n",
    "            \n",
    "            print('Accuracy: \\t{0:.2f}'.format(accuracy * 100.0))\n",
    "            print('Recall: \\t{0:.2f}'.format(recall * 100.0))\n",
    "            print('precision: \\t{0:.2f}'.format(precision * 100.0))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3f628e5a-26a7-4774-8cbc-9913bb044787",
   "metadata": {},
   "source": [
    "## initializing nets"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b6f69015-3e3c-4cf3-b0cf-4d8f3f4b5abb",
   "metadata": {},
   "source": [
    "### lbfgs -> Quasi Newton Verfahren\n",
    "<li> numerisches Verfahren </li>\n",
    "<li> nähern sich der Inversen der Hesse-Matrix an </li>\n",
    "<li> Rechenaufwandsminimierung pro Iteration </li>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "2a21184a-dd71-4064-adad-44372e811ed5",
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " This problem is unconstrained.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RUNNING THE L-BFGS-B CODE\n",
      "\n",
      "           * * *\n",
      "\n",
      "Machine precision = 2.220D-16\n",
      " N =         2087     M =           10\n",
      "\n",
      "At X0         0 variables are exactly at the bounds\n",
      "\n",
      "At iterate    0    f=  1.90888D+00    |proj g|=  2.92933D-01\n",
      "\n",
      "At iterate    1    f=  1.33391D+00    |proj g|=  7.69136D-02\n",
      "\n",
      "At iterate    2    f=  1.16951D+00    |proj g|=  4.38589D-02\n",
      "\n",
      "At iterate    3    f=  1.05721D+00    |proj g|=  1.39369D-01\n",
      "\n",
      "At iterate    4    f=  9.93131D-01    |proj g|=  1.70854D-01\n",
      "\n",
      "At iterate    5    f=  9.49129D-01    |proj g|=  5.02513D-02\n",
      "\n",
      "At iterate    6    f=  9.24464D-01    |proj g|=  5.96199D-02\n",
      "\n",
      "At iterate    7    f=  8.86298D-01    |proj g|=  1.02838D-01\n",
      "\n",
      "At iterate    8    f=  8.28345D-01    |proj g|=  2.84869D-02\n",
      "\n",
      "At iterate    9    f=  8.05274D-01    |proj g|=  5.89612D-02\n",
      "\n",
      "At iterate   10    f=  7.87550D-01    |proj g|=  2.48563D-02\n",
      "\n",
      "At iterate   11    f=  7.59983D-01    |proj g|=  4.31337D-02\n",
      "\n",
      "At iterate   12    f=  7.39788D-01    |proj g|=  2.38916D-02\n",
      "\n",
      "At iterate   13    f=  7.15261D-01    |proj g|=  3.36026D-02\n",
      "\n",
      "At iterate   14    f=  6.98147D-01    |proj g|=  3.94013D-02\n",
      "\n",
      "At iterate   15    f=  6.89261D-01    |proj g|=  2.90247D-02\n",
      "\n",
      "At iterate   16    f=  6.83690D-01    |proj g|=  2.24308D-02\n",
      "\n",
      "At iterate   17    f=  6.74234D-01    |proj g|=  2.13327D-02\n",
      "\n",
      "At iterate   18    f=  6.69539D-01    |proj g|=  2.89826D-02\n",
      "\n",
      "At iterate   19    f=  6.63545D-01    |proj g|=  1.09111D-02\n",
      "\n",
      "At iterate   20    f=  6.58582D-01    |proj g|=  1.27037D-02\n",
      "\n",
      "At iterate   21    f=  6.53556D-01    |proj g|=  3.41716D-02\n",
      "\n",
      "At iterate   22    f=  6.49547D-01    |proj g|=  3.06375D-02\n",
      "\n",
      "At iterate   23    f=  6.46476D-01    |proj g|=  1.13539D-02\n",
      "\n",
      "At iterate   24    f=  6.43987D-01    |proj g|=  1.34617D-02\n",
      "\n",
      "At iterate   25    f=  6.38965D-01    |proj g|=  2.19534D-02\n",
      "\n",
      "At iterate   26    f=  6.36509D-01    |proj g|=  1.12336D-01\n",
      "\n",
      "At iterate   27    f=  6.30912D-01    |proj g|=  2.22505D-02\n",
      "\n",
      "At iterate   28    f=  6.29908D-01    |proj g|=  8.80803D-03\n",
      "\n",
      "At iterate   29    f=  6.27144D-01    |proj g|=  8.78249D-03\n",
      "\n",
      "At iterate   30    f=  6.23719D-01    |proj g|=  5.34403D-02\n",
      "\n",
      "At iterate   31    f=  6.20344D-01    |proj g|=  2.12006D-02\n",
      "\n",
      "At iterate   32    f=  6.18876D-01    |proj g|=  7.21795D-03\n",
      "\n",
      "At iterate   33    f=  6.17289D-01    |proj g|=  1.26641D-02\n",
      "\n",
      "At iterate   34    f=  6.14821D-01    |proj g|=  5.31562D-02\n",
      "\n",
      "At iterate   35    f=  6.12148D-01    |proj g|=  1.99429D-02\n",
      "\n",
      "At iterate   36    f=  6.10594D-01    |proj g|=  1.04363D-02\n",
      "\n",
      "At iterate   37    f=  6.07669D-01    |proj g|=  2.14356D-02\n",
      "\n",
      "At iterate   38    f=  6.05002D-01    |proj g|=  2.07106D-02\n",
      "\n",
      "At iterate   39    f=  6.04592D-01    |proj g|=  9.90191D-02\n",
      "\n",
      "At iterate   40    f=  6.02124D-01    |proj g|=  2.24227D-02\n",
      "\n",
      "At iterate   41    f=  6.01384D-01    |proj g|=  6.41503D-03\n",
      "\n",
      "At iterate   42    f=  6.00417D-01    |proj g|=  2.96053D-02\n",
      "\n",
      "At iterate   43    f=  5.99083D-01    |proj g|=  3.58444D-02\n",
      "\n",
      "At iterate   44    f=  5.98613D-01    |proj g|=  8.07178D-02\n",
      "\n",
      "At iterate   45    f=  5.95371D-01    |proj g|=  2.31641D-02\n",
      "\n",
      "At iterate   46    f=  5.94254D-01    |proj g|=  8.07159D-03\n",
      "\n",
      "At iterate   47    f=  5.93134D-01    |proj g|=  2.77765D-02\n",
      "\n",
      "At iterate   48    f=  5.91952D-01    |proj g|=  2.42711D-02\n",
      "\n",
      "At iterate   49    f=  5.90396D-01    |proj g|=  1.45592D-02\n",
      "\n",
      "At iterate   50    f=  5.88096D-01    |proj g|=  1.36838D-02\n",
      "\n",
      "At iterate   51    f=  5.86164D-01    |proj g|=  2.40646D-02\n",
      "\n",
      "At iterate   52    f=  5.85303D-01    |proj g|=  3.50801D-02\n",
      "\n",
      "At iterate   53    f=  5.84268D-01    |proj g|=  2.14536D-02\n",
      "\n",
      "At iterate   54    f=  5.83062D-01    |proj g|=  5.53622D-03\n",
      "\n",
      "At iterate   55    f=  5.82092D-01    |proj g|=  8.71131D-03\n",
      "\n",
      "At iterate   56    f=  5.80209D-01    |proj g|=  1.41864D-02\n",
      "\n",
      "At iterate   57    f=  5.79668D-01    |proj g|=  8.13078D-02\n",
      "\n",
      "At iterate   58    f=  5.77080D-01    |proj g|=  1.08664D-02\n",
      "\n",
      "At iterate   59    f=  5.76181D-01    |proj g|=  1.10794D-02\n",
      "\n",
      "At iterate   60    f=  5.74899D-01    |proj g|=  2.12654D-02\n",
      "\n",
      "At iterate   61    f=  5.73722D-01    |proj g|=  4.31951D-02\n",
      "\n",
      "At iterate   62    f=  5.72312D-01    |proj g|=  1.76369D-02\n",
      "\n",
      "At iterate   63    f=  5.71502D-01    |proj g|=  6.11605D-03\n",
      "\n",
      "At iterate   64    f=  5.70627D-01    |proj g|=  1.43608D-02\n",
      "\n",
      "At iterate   65    f=  5.69771D-01    |proj g|=  5.30090D-02\n",
      "\n",
      "At iterate   66    f=  5.68676D-01    |proj g|=  2.53629D-02\n",
      "\n",
      "At iterate   67    f=  5.67385D-01    |proj g|=  8.57473D-03\n",
      "\n",
      "At iterate   68    f=  5.66714D-01    |proj g|=  1.15419D-02\n",
      "\n",
      "At iterate   69    f=  5.64152D-01    |proj g|=  9.18710D-03\n",
      "\n",
      "At iterate   70    f=  5.63183D-01    |proj g|=  4.37852D-02\n",
      "\n",
      "At iterate   71    f=  5.61854D-01    |proj g|=  9.04960D-03\n",
      "\n",
      "At iterate   72    f=  5.61078D-01    |proj g|=  8.91762D-03\n",
      "\n",
      "At iterate   73    f=  5.59611D-01    |proj g|=  1.48581D-02\n",
      "\n",
      "At iterate   74    f=  5.58998D-01    |proj g|=  4.40485D-02\n",
      "\n",
      "At iterate   75    f=  5.57714D-01    |proj g|=  8.96335D-03\n",
      "\n",
      "At iterate   76    f=  5.56979D-01    |proj g|=  7.98562D-03\n",
      "\n",
      "At iterate   77    f=  5.56344D-01    |proj g|=  1.27906D-02\n",
      "\n",
      "At iterate   78    f=  5.55799D-01    |proj g|=  5.66840D-02\n",
      "\n",
      "At iterate   79    f=  5.54809D-01    |proj g|=  7.53316D-03\n",
      "\n",
      "At iterate   80    f=  5.54323D-01    |proj g|=  1.06384D-02\n",
      "\n",
      "At iterate   81    f=  5.53517D-01    |proj g|=  1.38216D-02\n",
      "\n",
      "At iterate   82    f=  5.52612D-01    |proj g|=  2.97187D-02\n",
      "\n",
      "At iterate   83    f=  5.51464D-01    |proj g|=  1.24937D-02\n",
      "\n",
      "At iterate   84    f=  5.50584D-01    |proj g|=  7.94678D-03\n",
      "\n",
      "At iterate   85    f=  5.49218D-01    |proj g|=  7.40966D-03\n",
      "\n",
      "At iterate   86    f=  5.47836D-01    |proj g|=  2.08606D-02\n",
      "\n",
      "At iterate   87    f=  5.46435D-01    |proj g|=  1.51062D-02\n",
      "\n",
      "At iterate   88    f=  5.45588D-01    |proj g|=  1.48829D-02\n",
      "\n",
      "At iterate   89    f=  5.44892D-01    |proj g|=  8.58275D-03\n",
      "\n",
      "At iterate   90    f=  5.44197D-01    |proj g|=  8.02929D-03\n",
      "\n",
      "At iterate   91    f=  5.42437D-01    |proj g|=  2.86385D-02\n",
      "\n",
      "At iterate   92    f=  5.41988D-01    |proj g|=  2.65384D-02\n",
      "\n",
      "At iterate   93    f=  5.41232D-01    |proj g|=  1.26686D-02\n",
      "\n",
      "At iterate   94    f=  5.40497D-01    |proj g|=  1.30813D-02\n",
      "\n",
      "At iterate   95    f=  5.39988D-01    |proj g|=  4.61088D-02\n",
      "\n",
      "At iterate   96    f=  5.39085D-01    |proj g|=  1.55010D-02\n",
      "\n",
      "At iterate   97    f=  5.38388D-01    |proj g|=  8.64332D-03\n",
      "\n",
      "At iterate   98    f=  5.37898D-01    |proj g|=  1.23903D-02\n",
      "\n",
      "At iterate   99    f=  5.36896D-01    |proj g|=  1.24995D-02\n",
      "\n",
      "At iterate  100    f=  5.36418D-01    |proj g|=  4.13746D-02\n",
      "\n",
      "At iterate  101    f=  5.34852D-01    |proj g|=  1.73028D-02\n",
      "\n",
      "At iterate  102    f=  5.34301D-01    |proj g|=  9.47628D-03\n",
      "\n",
      "At iterate  103    f=  5.33782D-01    |proj g|=  3.96397D-02\n",
      "\n",
      "At iterate  104    f=  5.33257D-01    |proj g|=  1.19142D-02\n",
      "\n",
      "At iterate  105    f=  5.32806D-01    |proj g|=  1.40495D-02\n",
      "\n",
      "At iterate  106    f=  5.31815D-01    |proj g|=  1.88727D-02\n",
      "\n",
      "At iterate  107    f=  5.30862D-01    |proj g|=  1.30331D-02\n",
      "\n",
      "At iterate  108    f=  5.30608D-01    |proj g|=  4.62865D-02\n",
      "\n",
      "At iterate  109    f=  5.29592D-01    |proj g|=  8.47011D-03\n",
      "\n",
      "At iterate  110    f=  5.29377D-01    |proj g|=  1.17536D-02\n",
      "\n",
      "At iterate  111    f=  5.28657D-01    |proj g|=  1.83113D-02\n",
      "\n",
      "At iterate  112    f=  5.27385D-01    |proj g|=  2.20425D-02\n",
      "\n",
      "At iterate  113    f=  5.26810D-01    |proj g|=  2.98938D-02\n",
      "\n",
      "At iterate  114    f=  5.25441D-01    |proj g|=  2.03022D-02\n",
      "\n",
      "At iterate  115    f=  5.24982D-01    |proj g|=  4.10563D-02\n",
      "\n",
      "At iterate  116    f=  5.24503D-01    |proj g|=  2.26978D-02\n",
      "\n",
      "At iterate  117    f=  5.23842D-01    |proj g|=  6.36482D-03\n",
      "\n",
      "At iterate  118    f=  5.23133D-01    |proj g|=  1.18404D-02\n",
      "\n",
      "At iterate  119    f=  5.22084D-01    |proj g|=  1.53548D-02\n",
      "\n",
      "At iterate  120    f=  5.21480D-01    |proj g|=  3.71929D-02\n",
      "\n",
      "At iterate  121    f=  5.20360D-01    |proj g|=  1.20345D-02\n",
      "\n",
      "At iterate  122    f=  5.19550D-01    |proj g|=  1.49012D-02\n",
      "\n",
      "At iterate  123    f=  5.18924D-01    |proj g|=  2.36002D-02\n",
      "\n",
      "At iterate  124    f=  5.18488D-01    |proj g|=  1.43137D-02\n",
      "\n",
      "At iterate  125    f=  5.17781D-01    |proj g|=  7.14187D-03\n",
      "\n",
      "At iterate  126    f=  5.17312D-01    |proj g|=  4.32681D-02\n",
      "\n",
      "At iterate  127    f=  5.16722D-01    |proj g|=  2.81856D-02\n",
      "\n",
      "At iterate  128    f=  5.16027D-01    |proj g|=  2.28829D-02\n",
      "\n",
      "At iterate  129    f=  5.15466D-01    |proj g|=  2.49754D-02\n",
      "\n",
      "At iterate  130    f=  5.14976D-01    |proj g|=  2.18062D-02\n",
      "\n",
      "At iterate  131    f=  5.13763D-01    |proj g|=  1.17256D-02\n",
      "\n",
      "At iterate  132    f=  5.12450D-01    |proj g|=  1.36529D-02\n",
      "\n",
      "At iterate  133    f=  5.11919D-01    |proj g|=  2.26215D-02\n",
      "\n",
      "At iterate  134    f=  5.11155D-01    |proj g|=  1.44812D-02\n",
      "\n",
      "At iterate  135    f=  5.10385D-01    |proj g|=  8.68285D-03\n",
      "\n",
      "At iterate  136    f=  5.09919D-01    |proj g|=  1.55408D-02\n",
      "\n",
      "At iterate  137    f=  5.09672D-01    |proj g|=  4.38043D-02\n",
      "\n",
      "At iterate  138    f=  5.09326D-01    |proj g|=  1.75740D-02\n",
      "\n",
      "At iterate  139    f=  5.09005D-01    |proj g|=  8.08486D-03\n",
      "\n",
      "At iterate  140    f=  5.08736D-01    |proj g|=  1.66852D-02\n",
      "\n",
      "At iterate  141    f=  5.08243D-01    |proj g|=  1.90909D-02\n",
      "\n",
      "At iterate  142    f=  5.07474D-01    |proj g|=  4.42376D-02\n",
      "\n",
      "At iterate  143    f=  5.06815D-01    |proj g|=  2.56437D-02\n",
      "\n",
      "At iterate  144    f=  5.06062D-01    |proj g|=  2.02436D-02\n",
      "\n",
      "At iterate  145    f=  5.05311D-01    |proj g|=  2.57294D-02\n",
      "\n",
      "At iterate  146    f=  5.04981D-01    |proj g|=  2.04042D-02\n",
      "\n",
      "At iterate  147    f=  5.03970D-01    |proj g|=  1.30475D-02\n",
      "\n",
      "At iterate  148    f=  5.03037D-01    |proj g|=  2.33449D-02\n",
      "\n",
      "At iterate  149    f=  5.02470D-01    |proj g|=  4.56452D-02\n",
      "\n",
      "At iterate  150    f=  5.02074D-01    |proj g|=  8.19963D-03\n",
      "\n",
      "At iterate  151    f=  5.01959D-01    |proj g|=  9.62255D-03\n",
      "\n",
      "At iterate  152    f=  5.01581D-01    |proj g|=  1.61811D-02\n",
      "\n",
      "At iterate  153    f=  5.00880D-01    |proj g|=  1.72408D-02\n",
      "\n",
      "At iterate  154    f=  4.99881D-01    |proj g|=  2.34377D-02\n",
      "\n",
      "At iterate  155    f=  4.99284D-01    |proj g|=  1.08571D-02\n",
      "\n",
      "At iterate  156    f=  4.98729D-01    |proj g|=  7.32536D-03\n",
      "\n",
      "At iterate  157    f=  4.98214D-01    |proj g|=  2.11650D-02\n",
      "\n",
      "At iterate  158    f=  4.97763D-01    |proj g|=  1.46093D-02\n",
      "\n",
      "At iterate  159    f=  4.96999D-01    |proj g|=  8.26472D-03\n",
      "\n",
      "At iterate  160    f=  4.96671D-01    |proj g|=  3.83741D-02\n",
      "\n",
      "At iterate  161    f=  4.96158D-01    |proj g|=  1.04050D-02\n",
      "\n",
      "At iterate  162    f=  4.95842D-01    |proj g|=  8.09587D-03\n",
      "\n",
      "At iterate  163    f=  4.95085D-01    |proj g|=  4.16348D-02\n",
      "\n",
      "At iterate  164    f=  4.94490D-01    |proj g|=  8.94710D-03\n",
      "\n",
      "At iterate  165    f=  4.94160D-01    |proj g|=  1.69775D-02\n",
      "\n",
      "At iterate  166    f=  4.93827D-01    |proj g|=  2.09821D-02\n",
      "\n",
      "At iterate  167    f=  4.93237D-01    |proj g|=  1.31673D-02\n",
      "\n",
      "At iterate  168    f=  4.92855D-01    |proj g|=  2.73357D-02\n",
      "\n",
      "At iterate  169    f=  4.92284D-01    |proj g|=  1.20016D-02\n",
      "\n",
      "At iterate  170    f=  4.91836D-01    |proj g|=  1.83968D-02\n",
      "\n",
      "At iterate  171    f=  4.91332D-01    |proj g|=  1.99107D-02\n",
      "\n",
      "At iterate  172    f=  4.90927D-01    |proj g|=  4.53150D-02\n",
      "\n",
      "At iterate  173    f=  4.90240D-01    |proj g|=  1.35154D-02\n",
      "\n",
      "At iterate  174    f=  4.89882D-01    |proj g|=  8.60101D-03\n",
      "\n",
      "At iterate  175    f=  4.89157D-01    |proj g|=  1.18612D-02\n",
      "\n",
      "At iterate  176    f=  4.88919D-01    |proj g|=  2.44765D-02\n",
      "\n",
      "At iterate  177    f=  4.88467D-01    |proj g|=  2.37192D-02\n",
      "\n",
      "At iterate  178    f=  4.88099D-01    |proj g|=  7.85841D-03\n",
      "\n",
      "At iterate  179    f=  4.87774D-01    |proj g|=  9.53732D-03\n",
      "\n",
      "At iterate  180    f=  4.87470D-01    |proj g|=  3.93149D-02\n",
      "\n",
      "At iterate  181    f=  4.87021D-01    |proj g|=  9.23385D-03\n",
      "\n",
      "At iterate  182    f=  4.86667D-01    |proj g|=  1.08467D-02\n",
      "\n",
      "At iterate  183    f=  4.86275D-01    |proj g|=  1.84905D-02\n",
      "\n",
      "At iterate  184    f=  4.86072D-01    |proj g|=  5.41834D-02\n",
      "\n",
      "At iterate  185    f=  4.85610D-01    |proj g|=  1.13522D-02\n",
      "\n",
      "At iterate  186    f=  4.85364D-01    |proj g|=  7.05861D-03\n",
      "\n",
      "At iterate  187    f=  4.85016D-01    |proj g|=  1.40545D-02\n",
      "\n",
      "At iterate  188    f=  4.84441D-01    |proj g|=  1.68945D-02\n",
      "\n",
      "At iterate  189    f=  4.83742D-01    |proj g|=  2.05278D-02\n",
      "\n",
      "At iterate  190    f=  4.83436D-01    |proj g|=  4.09950D-02\n",
      "\n",
      "At iterate  191    f=  4.82711D-01    |proj g|=  1.21597D-02\n",
      "\n",
      "At iterate  192    f=  4.82404D-01    |proj g|=  6.99224D-03\n",
      "\n",
      "At iterate  193    f=  4.82145D-01    |proj g|=  2.31667D-02\n",
      "\n",
      "At iterate  194    f=  4.81836D-01    |proj g|=  8.90135D-03\n",
      "\n",
      "At iterate  195    f=  4.81418D-01    |proj g|=  5.90349D-03\n",
      "\n",
      "At iterate  196    f=  4.81084D-01    |proj g|=  1.98645D-02\n",
      "\n",
      "At iterate  197    f=  4.80687D-01    |proj g|=  1.09838D-02\n",
      "\n",
      "At iterate  198    f=  4.80322D-01    |proj g|=  9.50430D-03\n",
      "\n",
      "At iterate  199    f=  4.79912D-01    |proj g|=  6.40110D-03\n",
      "\n",
      "At iterate  200    f=  4.79366D-01    |proj g|=  8.94224D-03\n",
      "\n",
      "At iterate  201    f=  4.79131D-01    |proj g|=  3.62250D-02\n",
      "\n",
      "At iterate  202    f=  4.78391D-01    |proj g|=  8.30029D-03\n",
      "\n",
      "At iterate  203    f=  4.78117D-01    |proj g|=  5.89812D-03\n",
      "\n",
      "At iterate  204    f=  4.77751D-01    |proj g|=  1.93816D-02\n",
      "\n",
      "At iterate  205    f=  4.77497D-01    |proj g|=  3.10146D-02\n",
      "\n",
      "At iterate  206    f=  4.77165D-01    |proj g|=  1.05955D-02\n",
      "\n",
      "At iterate  207    f=  4.76889D-01    |proj g|=  1.17040D-02\n",
      "\n",
      "At iterate  208    f=  4.76658D-01    |proj g|=  1.18398D-02\n",
      "\n",
      "At iterate  209    f=  4.76119D-01    |proj g|=  1.64165D-02\n",
      "\n",
      "At iterate  210    f=  4.75843D-01    |proj g|=  2.19453D-02\n",
      "\n",
      "At iterate  211    f=  4.75465D-01    |proj g|=  1.82681D-02\n",
      "\n",
      "At iterate  212    f=  4.75237D-01    |proj g|=  3.66308D-02\n",
      "\n",
      "At iterate  213    f=  4.74967D-01    |proj g|=  1.14670D-02\n",
      "\n",
      "At iterate  214    f=  4.74719D-01    |proj g|=  6.57726D-03\n",
      "\n",
      "At iterate  215    f=  4.74410D-01    |proj g|=  1.49785D-02\n",
      "\n",
      "At iterate  216    f=  4.73850D-01    |proj g|=  2.22514D-02\n",
      "\n",
      "At iterate  217    f=  4.73711D-01    |proj g|=  1.43861D-02\n",
      "\n",
      "At iterate  218    f=  4.73308D-01    |proj g|=  7.34135D-03\n",
      "\n",
      "At iterate  219    f=  4.72919D-01    |proj g|=  6.77468D-03\n",
      "\n",
      "At iterate  220    f=  4.72451D-01    |proj g|=  2.02248D-02\n",
      "\n",
      "At iterate  221    f=  4.72093D-01    |proj g|=  1.46772D-02\n",
      "\n",
      "At iterate  222    f=  4.71852D-01    |proj g|=  7.31902D-03\n",
      "\n",
      "At iterate  223    f=  4.71440D-01    |proj g|=  1.19554D-02\n",
      "\n",
      "At iterate  224    f=  4.71062D-01    |proj g|=  1.88055D-02\n",
      "\n",
      "At iterate  225    f=  4.70692D-01    |proj g|=  1.61491D-02\n",
      "\n",
      "At iterate  226    f=  4.70396D-01    |proj g|=  1.51470D-02\n",
      "\n",
      "At iterate  227    f=  4.70187D-01    |proj g|=  7.05269D-03\n",
      "\n",
      "At iterate  228    f=  4.69818D-01    |proj g|=  9.11436D-03\n",
      "\n",
      "At iterate  229    f=  4.69690D-01    |proj g|=  2.56830D-02\n",
      "\n",
      "At iterate  230    f=  4.69416D-01    |proj g|=  1.42823D-02\n",
      "\n",
      "At iterate  231    f=  4.69144D-01    |proj g|=  5.76212D-03\n",
      "\n",
      "At iterate  232    f=  4.68858D-01    |proj g|=  7.73697D-03\n",
      "\n",
      "At iterate  233    f=  4.68487D-01    |proj g|=  1.23035D-02\n",
      "\n",
      "At iterate  234    f=  4.68109D-01    |proj g|=  3.51951D-02\n",
      "\n",
      "At iterate  235    f=  4.67692D-01    |proj g|=  1.29825D-02\n",
      "\n",
      "At iterate  236    f=  4.67488D-01    |proj g|=  5.58299D-03\n",
      "\n",
      "At iterate  237    f=  4.67259D-01    |proj g|=  9.82951D-03\n",
      "\n",
      "At iterate  238    f=  4.66995D-01    |proj g|=  3.13238D-02\n",
      "\n",
      "At iterate  239    f=  4.66624D-01    |proj g|=  1.29040D-02\n",
      "\n",
      "At iterate  240    f=  4.66319D-01    |proj g|=  2.76210D-02\n",
      "\n",
      "At iterate  241    f=  4.66064D-01    |proj g|=  1.79576D-02\n",
      "\n",
      "At iterate  242    f=  4.65901D-01    |proj g|=  1.61376D-02\n",
      "\n",
      "At iterate  243    f=  4.65403D-01    |proj g|=  8.85438D-03\n",
      "\n",
      "At iterate  244    f=  4.65071D-01    |proj g|=  1.05337D-02\n",
      "\n",
      "At iterate  245    f=  4.64812D-01    |proj g|=  6.69639D-03\n",
      "\n",
      "At iterate  246    f=  4.64431D-01    |proj g|=  8.56814D-03\n",
      "\n",
      "At iterate  247    f=  4.64325D-01    |proj g|=  3.64868D-02\n",
      "\n",
      "At iterate  248    f=  4.63972D-01    |proj g|=  8.54205D-03\n",
      "\n",
      "At iterate  249    f=  4.63724D-01    |proj g|=  7.23322D-03\n",
      "\n",
      "At iterate  250    f=  4.63458D-01    |proj g|=  1.00596D-02\n",
      "\n",
      "At iterate  251    f=  4.63040D-01    |proj g|=  2.09875D-02\n",
      "\n",
      "At iterate  252    f=  4.62838D-01    |proj g|=  2.50528D-02\n",
      "\n",
      "At iterate  253    f=  4.62583D-01    |proj g|=  9.06414D-03\n",
      "\n",
      "At iterate  254    f=  4.62338D-01    |proj g|=  6.88963D-03\n",
      "\n",
      "At iterate  255    f=  4.61871D-01    |proj g|=  1.10909D-02\n",
      "\n",
      "At iterate  256    f=  4.61499D-01    |proj g|=  2.43262D-02\n",
      "\n",
      "At iterate  257    f=  4.61136D-01    |proj g|=  1.38382D-02\n",
      "\n",
      "At iterate  258    f=  4.60907D-01    |proj g|=  7.28105D-03\n",
      "\n",
      "At iterate  259    f=  4.60649D-01    |proj g|=  8.64394D-03\n",
      "\n",
      "At iterate  260    f=  4.60401D-01    |proj g|=  1.30735D-02\n",
      "\n",
      "At iterate  261    f=  4.60171D-01    |proj g|=  3.09354D-02\n",
      "\n",
      "At iterate  262    f=  4.59851D-01    |proj g|=  1.04922D-02\n",
      "\n",
      "At iterate  263    f=  4.59729D-01    |proj g|=  8.99524D-03\n",
      "\n",
      "At iterate  264    f=  4.59561D-01    |proj g|=  2.18605D-02\n",
      "\n",
      "At iterate  265    f=  4.59331D-01    |proj g|=  2.10029D-02\n",
      "\n",
      "At iterate  266    f=  4.59133D-01    |proj g|=  2.25054D-02\n",
      "\n",
      "At iterate  267    f=  4.58669D-01    |proj g|=  1.25641D-02\n",
      "\n",
      "At iterate  268    f=  4.58605D-01    |proj g|=  3.57153D-02\n",
      "\n",
      "At iterate  269    f=  4.58251D-01    |proj g|=  1.39326D-02\n",
      "\n",
      "At iterate  270    f=  4.58045D-01    |proj g|=  9.03038D-03\n",
      "\n",
      "At iterate  271    f=  4.57787D-01    |proj g|=  1.15131D-02\n",
      "\n",
      "At iterate  272    f=  4.57411D-01    |proj g|=  1.18006D-02\n",
      "\n",
      "At iterate  273    f=  4.57151D-01    |proj g|=  3.20582D-02\n",
      "\n",
      "At iterate  274    f=  4.56783D-01    |proj g|=  1.09667D-02\n",
      "\n",
      "At iterate  275    f=  4.56612D-01    |proj g|=  7.09448D-03\n",
      "\n",
      "At iterate  276    f=  4.56278D-01    |proj g|=  1.38378D-02\n",
      "\n",
      "At iterate  277    f=  4.56123D-01    |proj g|=  2.60186D-02\n",
      "\n",
      "At iterate  278    f=  4.55847D-01    |proj g|=  1.35315D-02\n",
      "\n",
      "At iterate  279    f=  4.55538D-01    |proj g|=  6.79189D-03\n",
      "\n",
      "At iterate  280    f=  4.55111D-01    |proj g|=  1.09385D-02\n",
      "\n",
      "At iterate  281    f=  4.54850D-01    |proj g|=  2.40421D-02\n",
      "\n",
      "At iterate  282    f=  4.54436D-01    |proj g|=  7.55055D-03\n",
      "\n",
      "At iterate  283    f=  4.54217D-01    |proj g|=  5.78990D-03\n",
      "\n",
      "At iterate  284    f=  4.53945D-01    |proj g|=  9.50395D-03\n",
      "\n",
      "At iterate  285    f=  4.53804D-01    |proj g|=  3.63427D-02\n",
      "\n",
      "At iterate  286    f=  4.53474D-01    |proj g|=  1.21080D-02\n",
      "\n",
      "At iterate  287    f=  4.53299D-01    |proj g|=  1.10157D-02\n",
      "\n",
      "At iterate  288    f=  4.53089D-01    |proj g|=  1.09987D-02\n",
      "\n",
      "At iterate  289    f=  4.52782D-01    |proj g|=  3.42510D-02\n",
      "\n",
      "At iterate  290    f=  4.52453D-01    |proj g|=  1.17108D-02\n",
      "\n",
      "At iterate  291    f=  4.52207D-01    |proj g|=  7.75990D-03\n",
      "\n",
      "At iterate  292    f=  4.51922D-01    |proj g|=  8.68067D-03\n",
      "\n",
      "At iterate  293    f=  4.51604D-01    |proj g|=  1.36833D-02\n",
      "\n",
      "At iterate  294    f=  4.51405D-01    |proj g|=  2.77470D-02\n",
      "\n",
      "At iterate  295    f=  4.50938D-01    |proj g|=  7.19614D-03\n",
      "\n",
      "At iterate  296    f=  4.50794D-01    |proj g|=  8.74864D-03\n",
      "\n",
      "At iterate  297    f=  4.50324D-01    |proj g|=  1.87548D-02\n",
      "\n",
      "At iterate  298    f=  4.50151D-01    |proj g|=  2.35784D-02\n",
      "\n",
      "At iterate  299    f=  4.49932D-01    |proj g|=  9.05369D-03\n",
      "\n",
      "At iterate  300    f=  4.49697D-01    |proj g|=  5.86676D-03\n",
      "\n",
      "At iterate  301    f=  4.49311D-01    |proj g|=  1.15417D-02\n",
      "\n",
      "At iterate  302    f=  4.49163D-01    |proj g|=  2.61818D-02\n",
      "\n",
      "At iterate  303    f=  4.48934D-01    |proj g|=  1.11047D-02\n",
      "\n",
      "At iterate  304    f=  4.48705D-01    |proj g|=  5.38570D-03\n",
      "\n",
      "At iterate  305    f=  4.48496D-01    |proj g|=  8.79532D-03\n",
      "\n",
      "At iterate  306    f=  4.48132D-01    |proj g|=  1.26821D-02\n",
      "\n",
      "At iterate  307    f=  4.47813D-01    |proj g|=  2.79436D-02\n",
      "\n",
      "At iterate  308    f=  4.47335D-01    |proj g|=  9.58634D-03\n",
      "\n",
      "At iterate  309    f=  4.47098D-01    |proj g|=  9.72095D-03\n",
      "\n",
      "At iterate  310    f=  4.47018D-01    |proj g|=  1.69070D-02\n",
      "\n",
      "At iterate  311    f=  4.46878D-01    |proj g|=  1.27469D-02\n",
      "\n",
      "At iterate  312    f=  4.46669D-01    |proj g|=  7.65509D-03\n",
      "\n",
      "At iterate  313    f=  4.46291D-01    |proj g|=  1.30071D-02\n",
      "\n",
      "At iterate  314    f=  4.46072D-01    |proj g|=  2.16963D-02\n",
      "\n",
      "At iterate  315    f=  4.45837D-01    |proj g|=  1.38236D-02\n",
      "\n",
      "At iterate  316    f=  4.45553D-01    |proj g|=  1.07747D-02\n",
      "\n",
      "At iterate  317    f=  4.45368D-01    |proj g|=  2.89956D-02\n",
      "\n",
      "At iterate  318    f=  4.45209D-01    |proj g|=  1.76038D-02\n",
      "\n",
      "At iterate  319    f=  4.45010D-01    |proj g|=  6.04973D-03\n",
      "\n",
      "At iterate  320    f=  4.44881D-01    |proj g|=  1.30712D-02\n",
      "\n",
      "At iterate  321    f=  4.44588D-01    |proj g|=  1.95379D-02\n",
      "\n",
      "At iterate  322    f=  4.44298D-01    |proj g|=  2.57295D-02\n",
      "\n",
      "At iterate  323    f=  4.43953D-01    |proj g|=  9.64345D-03\n",
      "\n",
      "At iterate  324    f=  4.43727D-01    |proj g|=  6.63490D-03\n",
      "\n",
      "At iterate  325    f=  4.43508D-01    |proj g|=  3.51164D-02\n",
      "\n",
      "At iterate  326    f=  4.43328D-01    |proj g|=  1.15072D-02\n",
      "\n",
      "At iterate  327    f=  4.43211D-01    |proj g|=  5.33231D-03\n",
      "\n",
      "At iterate  328    f=  4.43031D-01    |proj g|=  1.24440D-02\n",
      "\n",
      "At iterate  329    f=  4.42888D-01    |proj g|=  7.87961D-03\n",
      "\n",
      "At iterate  330    f=  4.42636D-01    |proj g|=  1.25360D-02\n",
      "\n",
      "At iterate  331    f=  4.42294D-01    |proj g|=  7.62116D-03\n",
      "\n",
      "At iterate  332    f=  4.41943D-01    |proj g|=  5.63286D-03\n",
      "\n",
      "At iterate  333    f=  4.41776D-01    |proj g|=  2.30275D-02\n",
      "\n",
      "At iterate  334    f=  4.41530D-01    |proj g|=  9.09660D-03\n",
      "\n",
      "At iterate  335    f=  4.41348D-01    |proj g|=  8.10703D-03\n",
      "\n",
      "At iterate  336    f=  4.41135D-01    |proj g|=  1.16882D-02\n",
      "\n",
      "At iterate  337    f=  4.41031D-01    |proj g|=  2.58441D-02\n",
      "\n",
      "At iterate  338    f=  4.40883D-01    |proj g|=  6.67086D-03\n",
      "\n",
      "At iterate  339    f=  4.40771D-01    |proj g|=  6.92312D-03\n",
      "\n",
      "At iterate  340    f=  4.40599D-01    |proj g|=  1.12615D-02\n",
      "\n",
      "At iterate  341    f=  4.40239D-01    |proj g|=  1.30320D-02\n",
      "\n",
      "At iterate  342    f=  4.40013D-01    |proj g|=  1.77520D-02\n",
      "\n",
      "At iterate  343    f=  4.39649D-01    |proj g|=  8.86469D-03\n",
      "\n",
      "At iterate  344    f=  4.39471D-01    |proj g|=  2.48773D-02\n",
      "\n",
      "At iterate  345    f=  4.39276D-01    |proj g|=  7.36835D-03\n",
      "\n",
      "At iterate  346    f=  4.39114D-01    |proj g|=  7.07684D-03\n",
      "\n",
      "At iterate  347    f=  4.38866D-01    |proj g|=  1.15696D-02\n",
      "\n",
      "At iterate  348    f=  4.38712D-01    |proj g|=  2.70311D-02\n",
      "\n",
      "At iterate  349    f=  4.38507D-01    |proj g|=  2.65260D-02\n",
      "\n",
      "At iterate  350    f=  4.38361D-01    |proj g|=  4.79704D-03\n",
      "\n",
      "At iterate  351    f=  4.38303D-01    |proj g|=  5.37970D-03\n",
      "\n",
      "At iterate  352    f=  4.38153D-01    |proj g|=  7.63045D-03\n",
      "\n",
      "At iterate  353    f=  4.37798D-01    |proj g|=  1.52143D-02\n",
      "\n",
      "At iterate  354    f=  4.37629D-01    |proj g|=  2.00688D-02\n",
      "\n",
      "At iterate  355    f=  4.37445D-01    |proj g|=  7.64670D-03\n",
      "\n",
      "At iterate  356    f=  4.37299D-01    |proj g|=  8.52399D-03\n",
      "\n",
      "At iterate  357    f=  4.37125D-01    |proj g|=  9.82921D-03\n",
      "\n",
      "At iterate  358    f=  4.36861D-01    |proj g|=  3.01308D-02\n",
      "\n",
      "At iterate  359    f=  4.36649D-01    |proj g|=  1.98589D-02\n",
      "\n",
      "At iterate  360    f=  4.36464D-01    |proj g|=  9.55183D-03\n",
      "\n",
      "At iterate  361    f=  4.36399D-01    |proj g|=  4.72505D-02\n",
      "\n",
      "At iterate  362    f=  4.36221D-01    |proj g|=  1.68500D-02\n",
      "\n",
      "At iterate  363    f=  4.36137D-01    |proj g|=  5.88763D-03\n",
      "\n",
      "At iterate  364    f=  4.35954D-01    |proj g|=  1.57521D-02\n",
      "\n",
      "At iterate  365    f=  4.35831D-01    |proj g|=  1.43163D-02\n",
      "\n",
      "At iterate  366    f=  4.35669D-01    |proj g|=  2.38139D-02\n",
      "\n",
      "At iterate  367    f=  4.35404D-01    |proj g|=  1.27533D-02\n",
      "\n",
      "At iterate  368    f=  4.35205D-01    |proj g|=  9.44012D-03\n",
      "\n",
      "At iterate  369    f=  4.35039D-01    |proj g|=  8.25925D-03\n",
      "\n",
      "At iterate  370    f=  4.34817D-01    |proj g|=  9.76511D-03\n",
      "\n",
      "At iterate  371    f=  4.34659D-01    |proj g|=  1.62505D-02\n",
      "\n",
      "At iterate  372    f=  4.34462D-01    |proj g|=  9.19863D-03\n",
      "\n",
      "At iterate  373    f=  4.34334D-01    |proj g|=  1.12101D-02\n",
      "\n",
      "At iterate  374    f=  4.34157D-01    |proj g|=  1.67834D-02\n",
      "\n",
      "At iterate  375    f=  4.34081D-01    |proj g|=  2.34346D-02\n",
      "\n",
      "At iterate  376    f=  4.33948D-01    |proj g|=  8.97510D-03\n",
      "\n",
      "At iterate  377    f=  4.33841D-01    |proj g|=  4.87130D-03\n",
      "\n",
      "At iterate  378    f=  4.33610D-01    |proj g|=  6.47703D-03\n",
      "\n",
      "At iterate  379    f=  4.33386D-01    |proj g|=  1.59525D-02\n",
      "\n",
      "At iterate  380    f=  4.33158D-01    |proj g|=  6.78912D-03\n",
      "\n",
      "At iterate  381    f=  4.32958D-01    |proj g|=  6.54045D-03\n",
      "\n",
      "At iterate  382    f=  4.32738D-01    |proj g|=  1.22250D-02\n",
      "\n",
      "At iterate  383    f=  4.32591D-01    |proj g|=  1.86040D-02\n",
      "\n",
      "At iterate  384    f=  4.32441D-01    |proj g|=  1.09179D-02\n",
      "\n",
      "At iterate  385    f=  4.32205D-01    |proj g|=  9.11244D-03\n",
      "\n",
      "At iterate  386    f=  4.32157D-01    |proj g|=  2.10450D-02\n",
      "\n",
      "At iterate  387    f=  4.32072D-01    |proj g|=  1.40872D-02\n",
      "\n",
      "At iterate  388    f=  4.31917D-01    |proj g|=  5.56460D-03\n",
      "\n",
      "At iterate  389    f=  4.31761D-01    |proj g|=  9.40875D-03\n",
      "\n",
      "At iterate  390    f=  4.31551D-01    |proj g|=  1.02261D-02\n",
      "\n",
      "At iterate  391    f=  4.31525D-01    |proj g|=  6.30519D-02\n",
      "\n",
      "At iterate  392    f=  4.31240D-01    |proj g|=  1.69956D-02\n",
      "\n",
      "At iterate  393    f=  4.31139D-01    |proj g|=  5.01451D-03\n",
      "\n",
      "At iterate  394    f=  4.31036D-01    |proj g|=  1.24902D-02\n",
      "\n",
      "At iterate  395    f=  4.30915D-01    |proj g|=  8.67295D-03\n",
      "\n",
      "At iterate  396    f=  4.30873D-01    |proj g|=  2.16267D-02\n",
      "\n",
      "At iterate  397    f=  4.30633D-01    |proj g|=  2.17660D-02\n",
      "\n",
      "At iterate  398    f=  4.30534D-01    |proj g|=  1.36831D-02\n",
      "\n",
      "At iterate  399    f=  4.30465D-01    |proj g|=  6.16665D-03\n",
      "\n",
      "At iterate  400    f=  4.30358D-01    |proj g|=  8.43917D-03\n",
      "\n",
      "At iterate  401    f=  4.30242D-01    |proj g|=  9.51150D-03\n",
      "\n",
      "At iterate  402    f=  4.30052D-01    |proj g|=  1.35191D-02\n",
      "\n",
      "At iterate  403    f=  4.29891D-01    |proj g|=  8.29602D-03\n",
      "\n",
      "At iterate  404    f=  4.29780D-01    |proj g|=  5.09322D-03\n",
      "\n",
      "At iterate  405    f=  4.29719D-01    |proj g|=  3.02964D-02\n",
      "\n",
      "At iterate  406    f=  4.29596D-01    |proj g|=  6.38243D-03\n",
      "\n",
      "At iterate  407    f=  4.29548D-01    |proj g|=  6.56204D-03\n",
      "\n",
      "At iterate  408    f=  4.29393D-01    |proj g|=  9.07747D-03\n",
      "\n",
      "At iterate  409    f=  4.29303D-01    |proj g|=  1.83040D-02\n",
      "\n",
      "At iterate  410    f=  4.29185D-01    |proj g|=  8.65809D-03\n",
      "\n",
      "At iterate  411    f=  4.29086D-01    |proj g|=  4.87072D-03\n",
      "\n",
      "At iterate  412    f=  4.28943D-01    |proj g|=  4.85248D-03\n",
      "\n",
      "At iterate  413    f=  4.28733D-01    |proj g|=  1.48510D-02\n",
      "\n",
      "At iterate  414    f=  4.28544D-01    |proj g|=  5.13556D-03\n",
      "\n",
      "At iterate  415    f=  4.28394D-01    |proj g|=  5.46984D-03\n",
      "\n",
      "At iterate  416    f=  4.28206D-01    |proj g|=  9.99624D-03\n",
      "\n",
      "At iterate  417    f=  4.28088D-01    |proj g|=  1.30643D-02\n",
      "\n",
      "At iterate  418    f=  4.27984D-01    |proj g|=  2.03471D-02\n",
      "\n",
      "At iterate  419    f=  4.27874D-01    |proj g|=  1.19143D-02\n",
      "\n",
      "At iterate  420    f=  4.27746D-01    |proj g|=  6.04622D-03\n",
      "\n",
      "At iterate  421    f=  4.27633D-01    |proj g|=  5.36048D-03\n",
      "\n",
      "At iterate  422    f=  4.27534D-01    |proj g|=  3.04781D-02\n",
      "\n",
      "At iterate  423    f=  4.27355D-01    |proj g|=  1.15314D-02\n",
      "\n",
      "At iterate  424    f=  4.27273D-01    |proj g|=  4.89396D-03\n",
      "\n",
      "At iterate  425    f=  4.27178D-01    |proj g|=  5.88389D-03\n",
      "\n",
      "At iterate  426    f=  4.27058D-01    |proj g|=  8.67670D-03\n",
      "\n",
      "At iterate  427    f=  4.26847D-01    |proj g|=  8.65554D-03\n",
      "\n",
      "At iterate  428    f=  4.26761D-01    |proj g|=  1.54773D-02\n",
      "\n",
      "At iterate  429    f=  4.26640D-01    |proj g|=  4.97912D-03\n",
      "\n",
      "At iterate  430    f=  4.26539D-01    |proj g|=  8.46715D-03\n",
      "\n",
      "At iterate  431    f=  4.26478D-01    |proj g|=  2.18675D-02\n",
      "\n",
      "At iterate  432    f=  4.26398D-01    |proj g|=  1.60959D-02\n",
      "\n",
      "At iterate  433    f=  4.26177D-01    |proj g|=  4.92179D-03\n",
      "\n",
      "At iterate  434    f=  4.26087D-01    |proj g|=  1.16753D-02\n",
      "\n",
      "At iterate  435    f=  4.26016D-01    |proj g|=  1.65482D-02\n",
      "\n",
      "At iterate  436    f=  4.25929D-01    |proj g|=  1.14771D-02\n",
      "\n",
      "At iterate  437    f=  4.25772D-01    |proj g|=  8.35393D-03\n",
      "\n",
      "At iterate  438    f=  4.25678D-01    |proj g|=  8.10279D-03\n",
      "\n",
      "At iterate  439    f=  4.25475D-01    |proj g|=  8.13220D-03\n",
      "\n",
      "At iterate  440    f=  4.25368D-01    |proj g|=  1.75357D-02\n",
      "\n",
      "At iterate  441    f=  4.25208D-01    |proj g|=  8.64808D-03\n",
      "\n",
      "At iterate  442    f=  4.25091D-01    |proj g|=  8.14858D-03\n",
      "\n",
      "At iterate  443    f=  4.24980D-01    |proj g|=  8.22467D-03\n",
      "\n",
      "At iterate  444    f=  4.24891D-01    |proj g|=  3.74765D-02\n",
      "\n",
      "At iterate  445    f=  4.24769D-01    |proj g|=  1.91771D-02\n",
      "\n",
      "At iterate  446    f=  4.24631D-01    |proj g|=  9.13303D-03\n",
      "\n",
      "At iterate  447    f=  4.24571D-01    |proj g|=  5.62216D-03\n",
      "\n",
      "At iterate  448    f=  4.24519D-01    |proj g|=  3.19451D-02\n",
      "\n",
      "At iterate  449    f=  4.24342D-01    |proj g|=  9.16554D-03\n",
      "\n",
      "At iterate  450    f=  4.24291D-01    |proj g|=  9.44898D-03\n",
      "\n",
      "At iterate  451    f=  4.24125D-01    |proj g|=  9.33382D-03\n",
      "\n",
      "At iterate  452    f=  4.23868D-01    |proj g|=  6.75793D-03\n",
      "\n",
      "At iterate  453    f=  4.23761D-01    |proj g|=  2.27989D-02\n",
      "\n",
      "At iterate  454    f=  4.23588D-01    |proj g|=  6.63405D-03\n",
      "\n",
      "At iterate  455    f=  4.23492D-01    |proj g|=  4.95337D-03\n",
      "\n",
      "At iterate  456    f=  4.23357D-01    |proj g|=  8.59260D-03\n",
      "\n",
      "At iterate  457    f=  4.23245D-01    |proj g|=  2.87090D-02\n",
      "\n",
      "At iterate  458    f=  4.23137D-01    |proj g|=  1.99081D-02\n",
      "\n",
      "At iterate  459    f=  4.23029D-01    |proj g|=  8.48813D-03\n",
      "\n",
      "At iterate  460    f=  4.22947D-01    |proj g|=  5.77734D-03\n",
      "\n",
      "At iterate  461    f=  4.22886D-01    |proj g|=  6.67841D-03\n",
      "\n",
      "At iterate  462    f=  4.22835D-01    |proj g|=  1.88504D-02\n",
      "\n",
      "At iterate  463    f=  4.22756D-01    |proj g|=  7.50934D-03\n",
      "\n",
      "At iterate  464    f=  4.22689D-01    |proj g|=  5.77111D-03\n",
      "\n",
      "At iterate  465    f=  4.22625D-01    |proj g|=  1.07354D-02\n",
      "\n",
      "At iterate  466    f=  4.22495D-01    |proj g|=  1.27816D-02\n",
      "\n",
      "At iterate  467    f=  4.22281D-01    |proj g|=  1.50953D-02\n",
      "\n",
      "At iterate  468    f=  4.22153D-01    |proj g|=  1.71375D-02\n",
      "\n",
      "At iterate  469    f=  4.22013D-01    |proj g|=  9.47511D-03\n",
      "\n",
      "At iterate  470    f=  4.21932D-01    |proj g|=  6.41954D-03\n",
      "\n",
      "At iterate  471    f=  4.21864D-01    |proj g|=  2.00364D-02\n",
      "\n",
      "At iterate  472    f=  4.21783D-01    |proj g|=  8.93786D-03\n",
      "\n",
      "At iterate  473    f=  4.21633D-01    |proj g|=  1.09672D-02\n",
      "\n",
      "At iterate  474    f=  4.21574D-01    |proj g|=  8.97621D-03\n",
      "\n",
      "At iterate  475    f=  4.21509D-01    |proj g|=  2.14093D-02\n",
      "\n",
      "At iterate  476    f=  4.21409D-01    |proj g|=  6.60147D-03\n",
      "\n",
      "At iterate  477    f=  4.21345D-01    |proj g|=  8.02187D-03\n",
      "\n",
      "At iterate  478    f=  4.21272D-01    |proj g|=  9.27984D-03\n",
      "\n",
      "At iterate  479    f=  4.21108D-01    |proj g|=  9.84229D-03\n",
      "\n",
      "At iterate  480    f=  4.20955D-01    |proj g|=  1.71587D-02\n",
      "\n",
      "At iterate  481    f=  4.20815D-01    |proj g|=  1.22614D-02\n",
      "\n",
      "At iterate  482    f=  4.20751D-01    |proj g|=  4.88667D-03\n",
      "\n",
      "At iterate  483    f=  4.20656D-01    |proj g|=  5.50321D-03\n",
      "\n",
      "At iterate  484    f=  4.20568D-01    |proj g|=  1.84076D-02\n",
      "\n",
      "At iterate  485    f=  4.20450D-01    |proj g|=  9.85357D-03\n",
      "\n",
      "At iterate  486    f=  4.20346D-01    |proj g|=  1.34295D-02\n",
      "\n",
      "At iterate  487    f=  4.20305D-01    |proj g|=  1.24125D-02\n",
      "\n",
      "At iterate  488    f=  4.20273D-01    |proj g|=  1.05230D-02\n",
      "\n",
      "At iterate  489    f=  4.20181D-01    |proj g|=  6.93640D-03\n",
      "\n",
      "At iterate  490    f=  4.20093D-01    |proj g|=  7.67729D-03\n",
      "\n",
      "At iterate  491    f=  4.19949D-01    |proj g|=  7.05630D-03\n",
      "\n",
      "At iterate  492    f=  4.19842D-01    |proj g|=  5.48260D-03\n",
      "\n",
      "At iterate  493    f=  4.19750D-01    |proj g|=  1.09525D-02\n",
      "\n",
      "At iterate  494    f=  4.19661D-01    |proj g|=  5.67427D-03\n",
      "\n",
      "At iterate  495    f=  4.19543D-01    |proj g|=  5.65271D-03\n",
      "\n",
      "At iterate  496    f=  4.19459D-01    |proj g|=  9.76470D-03\n",
      "\n",
      "At iterate  497    f=  4.19348D-01    |proj g|=  7.44995D-03\n",
      "\n",
      "At iterate  498    f=  4.19252D-01    |proj g|=  7.51495D-03\n",
      "\n",
      "At iterate  499    f=  4.19216D-01    |proj g|=  1.35551D-02\n",
      "\n",
      "At iterate  500    f=  4.19157D-01    |proj g|=  4.80565D-03\n",
      "\n",
      "           * * *\n",
      "\n",
      "Tit   = total number of iterations\n",
      "Tnf   = total number of function evaluations\n",
      "Tnint = total number of segments explored during Cauchy searches\n",
      "Skip  = number of BFGS updates skipped\n",
      "Nact  = number of active bounds at final generalized Cauchy point\n",
      "Projg = norm of the final projected gradient\n",
      "F     = final function value\n",
      "\n",
      "           * * *\n",
      "\n",
      "   N    Tit     Tnf  Tnint  Skip  Nact     Projg        F\n",
      " 2087    500    538      1     0     0   4.806D-03   4.192D-01\n",
      "  F =  0.41915663982533991     \n",
      "\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT                 \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/jupyterhub/lib/python3.9/site-packages/sklearn/neural_network/_multilayer_perceptron.py:549: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "  self.n_iter_ = _check_optimize_result(\"lbfgs\", opt_res, self.max_iter)\n"
     ]
    }
   ],
   "source": [
    "mlp_1 = MLPClassifier(hidden_layer_sizes=(20,20,20), max_iter = 500, early_stopping = True, verbose = True, solver = \"lbfgs\")\n",
    "mlp_1train = mlp_1.fit(X_train_c,y_train_c)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "58f68a4d-d556-4939-8676-976e5f0e6e00",
   "metadata": {},
   "source": [
    "### Stochastic Gradient Descent (SDG)\n",
    "Berechnung: $\\lambda_{t+1} = \\lambda _t - \\eta\\nabla l(\\lambda)$\n",
    "<li> Suche den Gradienten der Funktion im Hinblick auf die weiteren Parameter/Atribute </li>\n",
    "<li> Random Initial Value für die Parameter </li>\n",
    "<li> Update der Parameter </li>\n",
    "<li> Berechnung der Schrittgröße : $stepsize = gradient * learningRate $ </li>\n",
    "<li> Berechnung der neuen Parameter : $ParameterNeu = ParameterAlt - stepsize $ </li>\n",
    "<li> Funktion hat ihr Minimum erreicht, wenn $ steigung = 0$ </li>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "d4f02a3f-d77b-4d44-81cb-b207614dd4e2",
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration 1, loss = 0.96145108\n",
      "Validation score: 0.701556\n",
      "Iteration 2, loss = 0.69272969\n",
      "Validation score: 0.729622\n",
      "Iteration 3, loss = 0.65079264\n",
      "Validation score: 0.733454\n",
      "Iteration 4, loss = 0.63049610\n",
      "Validation score: 0.738847\n",
      "Iteration 5, loss = 0.61658508\n",
      "Validation score: 0.743758\n",
      "Iteration 6, loss = 0.60526334\n",
      "Validation score: 0.746168\n",
      "Iteration 7, loss = 0.59506321\n",
      "Validation score: 0.751790\n",
      "Iteration 8, loss = 0.58617446\n",
      "Validation score: 0.755209\n",
      "Iteration 9, loss = 0.57874066\n",
      "Validation score: 0.758629\n",
      "Iteration 10, loss = 0.57211530\n",
      "Validation score: 0.761015\n",
      "Iteration 11, loss = 0.56622315\n",
      "Validation score: 0.762163\n",
      "Iteration 12, loss = 0.56089665\n",
      "Validation score: 0.766661\n",
      "Iteration 13, loss = 0.55604157\n",
      "Validation score: 0.767946\n",
      "Iteration 14, loss = 0.55117703\n",
      "Validation score: 0.768932\n",
      "Iteration 15, loss = 0.54696263\n",
      "Validation score: 0.770447\n",
      "Iteration 16, loss = 0.54283061\n",
      "Validation score: 0.772581\n",
      "Iteration 17, loss = 0.53849737\n",
      "Validation score: 0.774486\n",
      "Iteration 18, loss = 0.53419821\n",
      "Validation score: 0.776781\n",
      "Iteration 19, loss = 0.53057470\n",
      "Validation score: 0.777469\n",
      "Iteration 20, loss = 0.52661546\n",
      "Validation score: 0.779259\n",
      "Iteration 21, loss = 0.52282379\n",
      "Validation score: 0.782265\n",
      "Iteration 22, loss = 0.51908814\n",
      "Validation score: 0.783229\n",
      "Iteration 23, loss = 0.51550806\n",
      "Validation score: 0.784744\n",
      "Iteration 24, loss = 0.51201563\n",
      "Validation score: 0.786373\n",
      "Iteration 25, loss = 0.50902347\n",
      "Validation score: 0.786144\n",
      "Iteration 26, loss = 0.50620421\n",
      "Validation score: 0.786832\n",
      "Iteration 27, loss = 0.50322531\n",
      "Validation score: 0.788943\n",
      "Iteration 28, loss = 0.50055763\n",
      "Validation score: 0.789838\n",
      "Iteration 29, loss = 0.49763439\n",
      "Validation score: 0.793074\n",
      "Iteration 30, loss = 0.49501987\n",
      "Validation score: 0.791422\n",
      "Iteration 31, loss = 0.49247148\n",
      "Validation score: 0.792569\n",
      "Iteration 32, loss = 0.49026932\n",
      "Validation score: 0.792546\n",
      "Iteration 33, loss = 0.48802154\n",
      "Validation score: 0.794956\n",
      "Iteration 34, loss = 0.48578966\n",
      "Validation score: 0.797343\n",
      "Iteration 35, loss = 0.48402101\n",
      "Validation score: 0.798077\n",
      "Iteration 36, loss = 0.48241345\n",
      "Validation score: 0.793372\n",
      "Iteration 37, loss = 0.48057002\n",
      "Validation score: 0.799637\n",
      "Iteration 38, loss = 0.47899641\n",
      "Validation score: 0.797664\n",
      "Iteration 39, loss = 0.47727544\n",
      "Validation score: 0.799706\n",
      "Iteration 40, loss = 0.47558528\n",
      "Validation score: 0.801611\n",
      "Iteration 41, loss = 0.47415375\n",
      "Validation score: 0.803401\n",
      "Iteration 42, loss = 0.47242939\n",
      "Validation score: 0.803998\n",
      "Iteration 43, loss = 0.47109574\n",
      "Validation score: 0.799408\n",
      "Iteration 44, loss = 0.46976446\n",
      "Validation score: 0.803240\n",
      "Iteration 45, loss = 0.46842269\n",
      "Validation score: 0.802598\n",
      "Iteration 46, loss = 0.46708022\n",
      "Validation score: 0.805512\n",
      "Iteration 47, loss = 0.46574961\n",
      "Validation score: 0.806981\n",
      "Iteration 48, loss = 0.46456895\n",
      "Validation score: 0.797985\n",
      "Iteration 49, loss = 0.46340100\n",
      "Validation score: 0.805971\n",
      "Iteration 50, loss = 0.46238923\n",
      "Validation score: 0.804480\n",
      "Iteration 51, loss = 0.46129018\n",
      "Validation score: 0.808289\n",
      "Iteration 52, loss = 0.46029374\n",
      "Validation score: 0.805742\n",
      "Iteration 53, loss = 0.45916362\n",
      "Validation score: 0.807555\n",
      "Iteration 54, loss = 0.45820116\n",
      "Validation score: 0.808105\n",
      "Iteration 55, loss = 0.45750151\n",
      "Validation score: 0.808518\n",
      "Iteration 56, loss = 0.45647267\n",
      "Validation score: 0.810171\n",
      "Iteration 57, loss = 0.45544829\n",
      "Validation score: 0.808840\n",
      "Iteration 58, loss = 0.45428604\n",
      "Validation score: 0.809207\n",
      "Iteration 59, loss = 0.45353889\n",
      "Validation score: 0.811295\n",
      "Iteration 60, loss = 0.45239933\n",
      "Validation score: 0.810171\n",
      "Iteration 61, loss = 0.45192138\n",
      "Validation score: 0.811158\n",
      "Iteration 62, loss = 0.45075328\n",
      "Validation score: 0.813820\n",
      "Iteration 63, loss = 0.44998258\n",
      "Validation score: 0.812374\n",
      "Iteration 64, loss = 0.44908108\n",
      "Validation score: 0.812489\n",
      "Iteration 65, loss = 0.44816180\n",
      "Validation score: 0.812075\n",
      "Iteration 66, loss = 0.44722685\n",
      "Validation score: 0.813269\n",
      "Iteration 67, loss = 0.44662912\n",
      "Validation score: 0.814646\n",
      "Iteration 68, loss = 0.44563845\n",
      "Validation score: 0.814898\n",
      "Iteration 69, loss = 0.44489507\n",
      "Validation score: 0.817055\n",
      "Iteration 70, loss = 0.44417825\n",
      "Validation score: 0.816780\n",
      "Iteration 71, loss = 0.44347936\n",
      "Validation score: 0.816619\n",
      "Iteration 72, loss = 0.44272018\n",
      "Validation score: 0.816367\n",
      "Iteration 73, loss = 0.44216177\n",
      "Validation score: 0.817652\n",
      "Iteration 74, loss = 0.44142942\n",
      "Validation score: 0.816183\n",
      "Iteration 75, loss = 0.44091498\n",
      "Validation score: 0.817331\n",
      "Iteration 76, loss = 0.44018422\n",
      "Validation score: 0.817422\n",
      "Iteration 77, loss = 0.43938325\n",
      "Validation score: 0.817836\n",
      "Iteration 78, loss = 0.43881171\n",
      "Validation score: 0.820452\n",
      "Iteration 79, loss = 0.43819072\n",
      "Validation score: 0.818294\n",
      "Iteration 80, loss = 0.43781759\n",
      "Validation score: 0.818019\n",
      "Iteration 81, loss = 0.43700885\n",
      "Validation score: 0.817583\n",
      "Iteration 82, loss = 0.43641349\n",
      "Validation score: 0.820130\n",
      "Iteration 83, loss = 0.43586158\n",
      "Validation score: 0.816780\n",
      "Iteration 84, loss = 0.43496036\n",
      "Validation score: 0.817560\n",
      "Iteration 85, loss = 0.43439658\n",
      "Validation score: 0.820153\n",
      "Iteration 86, loss = 0.43401230\n",
      "Validation score: 0.816298\n",
      "Iteration 87, loss = 0.43361742\n",
      "Validation score: 0.822907\n",
      "Iteration 88, loss = 0.43267433\n",
      "Validation score: 0.817009\n",
      "Iteration 89, loss = 0.43234530\n",
      "Validation score: 0.821622\n",
      "Iteration 90, loss = 0.43133024\n",
      "Validation score: 0.821186\n",
      "Iteration 91, loss = 0.43069024\n",
      "Validation score: 0.823182\n",
      "Iteration 92, loss = 0.43014007\n",
      "Validation score: 0.820750\n",
      "Iteration 93, loss = 0.42953580\n",
      "Validation score: 0.824215\n",
      "Iteration 94, loss = 0.42909658\n",
      "Validation score: 0.822609\n",
      "Iteration 95, loss = 0.42818914\n",
      "Validation score: 0.823412\n",
      "Iteration 96, loss = 0.42748603\n",
      "Validation score: 0.824238\n",
      "Iteration 97, loss = 0.42683684\n",
      "Validation score: 0.825110\n",
      "Iteration 98, loss = 0.42638798\n",
      "Validation score: 0.823619\n",
      "Iteration 99, loss = 0.42561094\n",
      "Validation score: 0.826212\n",
      "Iteration 100, loss = 0.42522172\n",
      "Validation score: 0.825936\n",
      "Iteration 101, loss = 0.42435517\n",
      "Validation score: 0.821599\n",
      "Iteration 102, loss = 0.42442275\n",
      "Validation score: 0.825225\n",
      "Iteration 103, loss = 0.42382595\n",
      "Validation score: 0.826992\n",
      "Iteration 104, loss = 0.42323594\n",
      "Validation score: 0.824720\n",
      "Iteration 105, loss = 0.42278197\n",
      "Validation score: 0.825845\n",
      "Iteration 106, loss = 0.42246396\n",
      "Validation score: 0.827221\n",
      "Iteration 107, loss = 0.42206347\n",
      "Validation score: 0.826831\n",
      "Iteration 108, loss = 0.42177997\n",
      "Validation score: 0.826258\n",
      "Iteration 109, loss = 0.42146588\n",
      "Validation score: 0.826464\n",
      "Iteration 110, loss = 0.42089673\n",
      "Validation score: 0.828093\n",
      "Iteration 111, loss = 0.42080568\n",
      "Validation score: 0.828369\n",
      "Iteration 112, loss = 0.42081629\n",
      "Validation score: 0.826326\n",
      "Iteration 113, loss = 0.41996968\n",
      "Validation score: 0.822310\n",
      "Iteration 114, loss = 0.41963411\n",
      "Validation score: 0.825225\n",
      "Iteration 115, loss = 0.41896072\n",
      "Validation score: 0.825294\n",
      "Iteration 116, loss = 0.41907865\n",
      "Validation score: 0.829493\n",
      "Iteration 117, loss = 0.41858554\n",
      "Validation score: 0.825936\n",
      "Iteration 118, loss = 0.41852388\n",
      "Validation score: 0.827382\n",
      "Iteration 119, loss = 0.41814343\n",
      "Validation score: 0.825684\n",
      "Iteration 120, loss = 0.41771154\n",
      "Validation score: 0.826441\n",
      "Iteration 121, loss = 0.41738788\n",
      "Validation score: 0.824835\n",
      "Iteration 122, loss = 0.41734155\n",
      "Validation score: 0.821783\n",
      "Iteration 123, loss = 0.41720034\n",
      "Validation score: 0.827589\n",
      "Iteration 124, loss = 0.41659874\n",
      "Validation score: 0.828529\n",
      "Iteration 125, loss = 0.41640500\n",
      "Validation score: 0.823068\n",
      "Iteration 126, loss = 0.41627210\n",
      "Validation score: 0.828438\n",
      "Iteration 127, loss = 0.41619050\n",
      "Validation score: 0.827634\n",
      "Validation score did not improve more than tol=0.000100 for 10 consecutive epochs. Stopping.\n"
     ]
    }
   ],
   "source": [
    "mlp_2 = MLPClassifier(hidden_layer_sizes=(20,20,20),  max_iter = 150, early_stopping = True, verbose = True, solver = \"sgd\")\n",
    "mlp_2train = mlp_2.fit(X_train_c,y_train_c)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "65a6b755-46ba-48b7-a6c1-f46685c1e571",
   "metadata": {},
   "source": [
    "### Adam\n",
    "<p> Kombination aus zwei Erweiterungen </p>\n",
    "<li> Adaptive Gradient Algorithm </li>\n",
    "<li> Root Mean Square Propagation </li>\n",
    "\n",
    "#### Parameter\n",
    "<li> alpha: Schrittgröße/Lernrate </li>\n",
    "<li> beta1: Exponetielle Zerfallsrate des ersten Impulses </li>\n",
    "<li> beta2: Exponetielle Zerfallsrate des zweiten Impulses </li>\n",
    "<li> epsilon: Um nicht durch 0 zu teilen </li>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "e0a4874c-1d2a-4cd8-bce9-17f0a2a68efd",
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration 1, loss = 0.68355235\n",
      "Validation score: 0.749059\n",
      "Iteration 2, loss = 0.56364305\n",
      "Validation score: 0.766546\n",
      "Iteration 3, loss = 0.53554583\n",
      "Validation score: 0.775863\n",
      "Iteration 4, loss = 0.51575188\n",
      "Validation score: 0.783619\n",
      "Iteration 5, loss = 0.50174390\n",
      "Validation score: 0.787337\n",
      "Iteration 6, loss = 0.49190140\n",
      "Validation score: 0.794038\n",
      "Iteration 7, loss = 0.48325734\n",
      "Validation score: 0.796930\n",
      "Iteration 8, loss = 0.47729384\n",
      "Validation score: 0.796241\n",
      "Iteration 9, loss = 0.47120280\n",
      "Validation score: 0.797939\n",
      "Iteration 10, loss = 0.46726100\n",
      "Validation score: 0.801978\n",
      "Iteration 11, loss = 0.46283348\n",
      "Validation score: 0.800785\n",
      "Iteration 12, loss = 0.46042477\n",
      "Validation score: 0.807968\n",
      "Iteration 13, loss = 0.45783720\n",
      "Validation score: 0.798628\n",
      "Iteration 14, loss = 0.45527178\n",
      "Validation score: 0.807669\n",
      "Iteration 15, loss = 0.45290624\n",
      "Validation score: 0.809161\n",
      "Iteration 16, loss = 0.45139818\n",
      "Validation score: 0.807440\n",
      "Iteration 17, loss = 0.44938661\n",
      "Validation score: 0.811433\n",
      "Iteration 18, loss = 0.44676202\n",
      "Validation score: 0.809758\n",
      "Iteration 19, loss = 0.44574021\n",
      "Validation score: 0.809069\n",
      "Iteration 20, loss = 0.44384465\n",
      "Validation score: 0.814141\n",
      "Iteration 21, loss = 0.44146022\n",
      "Validation score: 0.815655\n",
      "Iteration 22, loss = 0.44097252\n",
      "Validation score: 0.813315\n",
      "Iteration 23, loss = 0.43909857\n",
      "Validation score: 0.809345\n",
      "Iteration 24, loss = 0.43824465\n",
      "Validation score: 0.811203\n",
      "Iteration 25, loss = 0.43713094\n",
      "Validation score: 0.815564\n",
      "Iteration 26, loss = 0.43556954\n",
      "Validation score: 0.815105\n",
      "Iteration 27, loss = 0.43444259\n",
      "Validation score: 0.818226\n",
      "Iteration 28, loss = 0.43404761\n",
      "Validation score: 0.820360\n",
      "Iteration 29, loss = 0.43216117\n",
      "Validation score: 0.817881\n",
      "Iteration 30, loss = 0.43151285\n",
      "Validation score: 0.814508\n",
      "Iteration 31, loss = 0.43022808\n",
      "Validation score: 0.819557\n",
      "Iteration 32, loss = 0.42928883\n",
      "Validation score: 0.820039\n",
      "Iteration 33, loss = 0.42901476\n",
      "Validation score: 0.814806\n",
      "Iteration 34, loss = 0.42733880\n",
      "Validation score: 0.817262\n",
      "Iteration 35, loss = 0.42717702\n",
      "Validation score: 0.822632\n",
      "Iteration 36, loss = 0.42653329\n",
      "Validation score: 0.817055\n",
      "Iteration 37, loss = 0.42502430\n",
      "Validation score: 0.818317\n",
      "Iteration 38, loss = 0.42459307\n",
      "Validation score: 0.821668\n",
      "Iteration 39, loss = 0.42376874\n",
      "Validation score: 0.822333\n",
      "Iteration 40, loss = 0.42316028\n",
      "Validation score: 0.821874\n",
      "Iteration 41, loss = 0.42215547\n",
      "Validation score: 0.823320\n",
      "Iteration 42, loss = 0.42150751\n",
      "Validation score: 0.824123\n",
      "Iteration 43, loss = 0.42113359\n",
      "Validation score: 0.821966\n",
      "Iteration 44, loss = 0.41982103\n",
      "Validation score: 0.822999\n",
      "Iteration 45, loss = 0.41953628\n",
      "Validation score: 0.821507\n",
      "Iteration 46, loss = 0.41858993\n",
      "Validation score: 0.823343\n",
      "Iteration 47, loss = 0.41794385\n",
      "Validation score: 0.825822\n",
      "Iteration 48, loss = 0.41776579\n",
      "Validation score: 0.822586\n",
      "Iteration 49, loss = 0.41685027\n",
      "Validation score: 0.828139\n",
      "Iteration 50, loss = 0.41595879\n",
      "Validation score: 0.824582\n",
      "Iteration 51, loss = 0.41542136\n",
      "Validation score: 0.825408\n",
      "Iteration 52, loss = 0.41508714\n",
      "Validation score: 0.826785\n",
      "Iteration 53, loss = 0.41433369\n",
      "Validation score: 0.824491\n",
      "Iteration 54, loss = 0.41328683\n",
      "Validation score: 0.829126\n",
      "Iteration 55, loss = 0.41328943\n",
      "Validation score: 0.826717\n",
      "Iteration 56, loss = 0.41227024\n",
      "Validation score: 0.826303\n",
      "Iteration 57, loss = 0.41141645\n",
      "Validation score: 0.826441\n",
      "Iteration 58, loss = 0.41132096\n",
      "Validation score: 0.827795\n",
      "Iteration 59, loss = 0.41070458\n",
      "Validation score: 0.826671\n",
      "Iteration 60, loss = 0.41074427\n",
      "Validation score: 0.823435\n",
      "Iteration 61, loss = 0.40939701\n",
      "Validation score: 0.829447\n",
      "Iteration 62, loss = 0.40993643\n",
      "Validation score: 0.828300\n",
      "Iteration 63, loss = 0.40883196\n",
      "Validation score: 0.828713\n",
      "Iteration 64, loss = 0.40901836\n",
      "Validation score: 0.829241\n",
      "Iteration 65, loss = 0.40820685\n",
      "Validation score: 0.827795\n",
      "Iteration 66, loss = 0.40868739\n",
      "Validation score: 0.830228\n",
      "Iteration 67, loss = 0.40838587\n",
      "Validation score: 0.824169\n",
      "Iteration 68, loss = 0.40748286\n",
      "Validation score: 0.827359\n",
      "Iteration 69, loss = 0.40758269\n",
      "Validation score: 0.828667\n",
      "Iteration 70, loss = 0.40718536\n",
      "Validation score: 0.825225\n",
      "Iteration 71, loss = 0.40640388\n",
      "Validation score: 0.829700\n",
      "Iteration 72, loss = 0.40684328\n",
      "Validation score: 0.826326\n",
      "Iteration 73, loss = 0.40643303\n",
      "Validation score: 0.829815\n",
      "Iteration 74, loss = 0.40631551\n",
      "Validation score: 0.825936\n",
      "Iteration 75, loss = 0.40627319\n",
      "Validation score: 0.828736\n",
      "Iteration 76, loss = 0.40579672\n",
      "Validation score: 0.827979\n",
      "Iteration 77, loss = 0.40509173\n",
      "Validation score: 0.828461\n",
      "Validation score did not improve more than tol=0.000100 for 10 consecutive epochs. Stopping.\n"
     ]
    }
   ],
   "source": [
    "mlp_3 = MLPClassifier(hidden_layer_sizes=(20,20,20), max_iter = 150, early_stopping = True, verbose = True,solver = \"adam\")\n",
    "mlp_3train = mlp_3.fit(X_train_c,y_train_c)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f4b323c8-2457-480e-958d-2acc38eec75b",
   "metadata": {},
   "source": [
    "__________________________________________________________________________________________________________________________________"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "4ac5986c-d3f6-4858-8305-52a9e57b4518",
   "metadata": {},
   "outputs": [],
   "source": [
    "addToList(mlp_1train, classificationList)\n",
    "addToList(mlp_2train, classificationList)\n",
    "addToList(mlp_3train, classificationList)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "56044537-957c-4290-a83a-73c02e499a82",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[MLPClassifier(early_stopping=True, hidden_layer_sizes=(20, 20, 20),\n",
       "               max_iter=500, solver='lbfgs', verbose=True),\n",
       " MLPClassifier(early_stopping=True, hidden_layer_sizes=(20, 20, 20),\n",
       "               max_iter=150, solver='sgd', verbose=True),\n",
       " MLPClassifier(early_stopping=True, hidden_layer_sizes=(20, 20, 20),\n",
       "               max_iter=150, verbose=True)]"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "classificationList"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6d98e013-6397-47bf-9663-2415ad6c92f8",
   "metadata": {},
   "source": [
    "## Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "2ad41f1b-5588-42f7-8797-cd2f50093c01",
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MLPClassifier(early_stopping=True, hidden_layer_sizes=(20, 20, 20),\n",
      "              max_iter=500, solver='lbfgs', verbose=True)\n",
      "\n",
      "paramters: {'activation': 'relu', 'alpha': 0.0001, 'batch_size': 'auto', 'beta_1': 0.9, 'beta_2': 0.999, 'early_stopping': True, 'epsilon': 1e-08, 'hidden_layer_sizes': (20, 20, 20), 'learning_rate': 'constant', 'learning_rate_init': 0.001, 'max_fun': 15000, 'max_iter': 500, 'momentum': 0.9, 'n_iter_no_change': 10, 'nesterovs_momentum': True, 'power_t': 0.5, 'random_state': None, 'shuffle': True, 'solver': 'lbfgs', 'tol': 0.0001, 'validation_fraction': 0.1, 'verbose': True, 'warm_start': False}\n",
      "\n",
      "last loss: 4.191566398253399e-01\n",
      "\n",
      "number of features seen during fit: 54\n",
      "\n",
      "number of layers: 5\n",
      "\n",
      "number of outputs: 7\n",
      "\n",
      "name of output activation function: softmax\n",
      "\n",
      "number of iterations: 500\n",
      "\n",
      "number training samples seen by the solver during fitting: 0\n",
      "****************************************************************************************\n",
      "MLPClassifier(early_stopping=True, hidden_layer_sizes=(20, 20, 20),\n",
      "              max_iter=150, solver='sgd', verbose=True)\n",
      "\n",
      "paramters: {'activation': 'relu', 'alpha': 0.0001, 'batch_size': 'auto', 'beta_1': 0.9, 'beta_2': 0.999, 'early_stopping': True, 'epsilon': 1e-08, 'hidden_layer_sizes': (20, 20, 20), 'learning_rate': 'constant', 'learning_rate_init': 0.001, 'max_fun': 15000, 'max_iter': 150, 'momentum': 0.9, 'n_iter_no_change': 10, 'nesterovs_momentum': True, 'power_t': 0.5, 'random_state': None, 'shuffle': True, 'solver': 'sgd', 'tol': 0.0001, 'validation_fraction': 0.1, 'verbose': True, 'warm_start': False}\n",
      "\n",
      "last loss: 0.41619050336143043\n",
      "\n",
      "number of features seen during fit: 54\n",
      "\n",
      "number of layers: 5\n",
      "\n",
      "number of outputs: 7\n",
      "\n",
      "name of output activation function: softmax\n",
      "\n",
      "number of iterations: 127\n",
      "\n",
      "number training samples seen by the solver during fitting: 49807241\n",
      "****************************************************************************************\n",
      "MLPClassifier(early_stopping=True, hidden_layer_sizes=(20, 20, 20),\n",
      "              max_iter=150, verbose=True)\n",
      "\n",
      "paramters: {'activation': 'relu', 'alpha': 0.0001, 'batch_size': 'auto', 'beta_1': 0.9, 'beta_2': 0.999, 'early_stopping': True, 'epsilon': 1e-08, 'hidden_layer_sizes': (20, 20, 20), 'learning_rate': 'constant', 'learning_rate_init': 0.001, 'max_fun': 15000, 'max_iter': 150, 'momentum': 0.9, 'n_iter_no_change': 10, 'nesterovs_momentum': True, 'power_t': 0.5, 'random_state': None, 'shuffle': True, 'solver': 'adam', 'tol': 0.0001, 'validation_fraction': 0.1, 'verbose': True, 'warm_start': False}\n",
      "\n",
      "last loss: 0.40509172520600584\n",
      "\n",
      "number of features seen during fit: 54\n",
      "\n",
      "number of layers: 5\n",
      "\n",
      "number of outputs: 7\n",
      "\n",
      "name of output activation function: softmax\n",
      "\n",
      "number of iterations: 77\n",
      "\n",
      "number training samples seen by the solver during fitting: 30198091\n",
      "****************************************************************************************\n"
     ]
    }
   ],
   "source": [
    "comparison(classificationList)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "daad5cc5-0f3e-416d-bf86-f53792ab7ba2",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "default: hidden_layer = (100,), activation = relu, solver = adam, alpha = 0.0001, learning_rate = constant, learning_rate_init = 0.001, max_iter = 200, shuffle = true, early_stopping & warm_start = FALSE\n",
      "Alle Infos, die davon abweichen, werden in der Legende gelistet\n",
      "****************************************************************************************\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# lbfgs kann keine Verlustfunktion plotten\n",
    "removeFromList(mlp_1train, classificationList)\n",
    "\n",
    "curveValidateIteration(classificationList, curve, validation, iteration)\n",
    "plotCurvePerInteration(classificationList, curve, iteration)\n",
    "plotValidationPerInteration(classificationList, validation, iteration)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "715808d2-3450-4561-b72a-fe134258b439",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plotValidationPercurve(classificationList, validation, curve)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cc09c0c7-abb7-44d0-bbe0-36da87b460e4",
   "metadata": {},
   "source": [
    "## Heatmaps"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "ef05d5a2-ac58-4035-aa6c-dfa6abd61000",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MLPClassifier(early_stopping=True, hidden_layer_sizes=(20, 20, 20),\n",
      "              max_iter=500, solver='lbfgs', verbose=True)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MLPClassifier(early_stopping=True, hidden_layer_sizes=(20, 20, 20),\n",
      "              max_iter=150, solver='sgd', verbose=True)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MLPClassifier(early_stopping=True, hidden_layer_sizes=(20, 20, 20),\n",
      "              max_iter=150, verbose=True)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "heatmaping(classificationList)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "6f083132-b371-422e-9d69-2489a9f2c6db",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MLPClassifier(early_stopping=True, hidden_layer_sizes=(20, 20, 20),\n",
      "              max_iter=500, solver='lbfgs', verbose=True)\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           1       0.84      0.81      0.83     52953\n",
      "           2       0.84      0.89      0.86     70876\n",
      "           3       0.74      0.83      0.78      8894\n",
      "           4       0.69      0.49      0.57       685\n",
      "           5       0.74      0.33      0.46      2323\n",
      "           6       0.59      0.42      0.49      4423\n",
      "           7       0.85      0.81      0.83      5099\n",
      "\n",
      "    accuracy                           0.83    145253\n",
      "   macro avg       0.76      0.65      0.69    145253\n",
      "weighted avg       0.82      0.83      0.82    145253\n",
      "\n",
      "MLPClassifier(early_stopping=True, hidden_layer_sizes=(20, 20, 20),\n",
      "              max_iter=150, solver='sgd', verbose=True)\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           1       0.84      0.81      0.83     52953\n",
      "           2       0.84      0.89      0.86     70876\n",
      "           3       0.76      0.83      0.79      8894\n",
      "           4       0.79      0.51      0.62       685\n",
      "           5       0.70      0.34      0.46      2323\n",
      "           6       0.60      0.53      0.56      4423\n",
      "           7       0.87      0.76      0.81      5099\n",
      "\n",
      "    accuracy                           0.83    145253\n",
      "   macro avg       0.77      0.67      0.70    145253\n",
      "weighted avg       0.83      0.83      0.83    145253\n",
      "\n",
      "MLPClassifier(early_stopping=True, hidden_layer_sizes=(20, 20, 20),\n",
      "              max_iter=150, verbose=True)\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           1       0.84      0.81      0.82     52953\n",
      "           2       0.84      0.89      0.86     70876\n",
      "           3       0.78      0.84      0.80      8894\n",
      "           4       0.83      0.57      0.68       685\n",
      "           5       0.69      0.44      0.54      2323\n",
      "           6       0.66      0.54      0.59      4423\n",
      "           7       0.87      0.79      0.83      5099\n",
      "\n",
      "    accuracy                           0.83    145253\n",
      "   macro avg       0.79      0.69      0.73    145253\n",
      "weighted avg       0.83      0.83      0.83    145253\n",
      "\n"
     ]
    }
   ],
   "source": [
    "reporting(classificationList)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "74a399f7-9f58-4c18-9122-6cd52a869c2f",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MLPClassifier(early_stopping=True, hidden_layer_sizes=(20, 20, 20),\n",
      "              max_iter=500, solver='lbfgs', verbose=True)\n",
      "Accuracy: \t82.66\n",
      "Recall: \t82.66\n",
      "precision: \t82.43\n",
      "MLPClassifier(early_stopping=True, hidden_layer_sizes=(20, 20, 20),\n",
      "              max_iter=150, solver='sgd', verbose=True)\n",
      "Accuracy: \t82.87\n",
      "Recall: \t82.87\n",
      "precision: \t82.73\n",
      "MLPClassifier(early_stopping=True, hidden_layer_sizes=(20, 20, 20),\n",
      "              max_iter=150, verbose=True)\n",
      "Accuracy: \t83.13\n",
      "Recall: \t83.13\n",
      "precision: \t82.98\n"
     ]
    }
   ],
   "source": [
    "accuracyPlotting(classificationList)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "03fabf70-d9f4-4968-902f-67dca632de9b",
   "metadata": {
    "tags": []
   },
   "source": [
    "## Fazit\n",
    "<li> Adam erzieht die besten Ergebnisse </li>\n",
    "<li> lbfgs konnte auch mit einer maximalen iteration von 500 nicht durchlaufen, ist aber dafür besonders schnell </li>\n",
    "<li> sdg beginnt mit einer deutlich niedrigen Verlustkurve, endet aber frühzeitiger, sodass Adam die besseren Ergebnisse vorweisen kann </li>"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}