IntroToML1.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyMqu33o7ENuqV9184oHdRsR",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/rukundob451/MachineLearning/blob/main/IntroToML1.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Supervised Learning and K Nearest Neighbors Exercises\n",
        "\n",
        "### Learning Objectives:\n",
        "\n",
        "* Explain supervised learning and how it can be applied to regression and classification problems\n",
        "* Apply K-Nearest Neighbor (KNN) algorithm for classification\n",
        "* Apply Intel® Extension for Scikit-learn* to leverage underlying compute capabilities of hardware\n",
        "\n",
        "\n",
        "\n"
      ],
      "metadata": {
        "id": "QerEvuSguirh"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### scikit-learn*\n",
        "\n",
        "Frameworks provide structure that Data Scientists use to build code. Frameworks are more than just libraries, because in addition to callable code, frameworks influence how code is written.\n",
        "\n",
        "A main virtue of using an optimized framework is that code runs faster. Code that runs faster is just generally more convenient but when we begin looking at applied data science and AI models, we can see more material benefits. Here you will see how optimization, particularly hyperparameter optimization can benefit more than just speed."
      ],
      "metadata": {
        "id": "OMscWeG6xgJA"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Introduction (Class 2)\n",
        "\n",
        "We will be using customer churn data from the telecom industry for the first week's exercises. The data file is called Orange_Telecom_Churn_Data.csv. We will load this data together, do some preprocessing, and use K-nearest neighbors to predict customer churn based on account characteristics."
      ],
      "metadata": {
        "id": "AXv56gg-zOgB"
      }
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "id": "ahxPJOl3s_R9"
      },
      "outputs": [],
      "source": [
        "from sklearn.preprocessing import LabelBinarizer\n",
        "from sklearn.preprocessing import MinMaxScaler\n",
        "from sklearn.neighbors import KNeighborsClassifier"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Question 1\n",
        "\n",
        "Begin by importing the data. Examine the columns and data.\n",
        "Notice that the data contains a state, area code, and phone number.\n",
        "Do you think these are good features to use when building a machine learning model? Why or why not?\n",
        "We will not be using them, so they can be dropped from the data."
      ],
      "metadata": {
        "id": "6A4yjX58zZNY"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "\n",
        "data = pd.read_csv('Orange_Telecom_Churn_Data.csv')"
      ],
      "metadata": {
        "id": "r_1Q1yixzILA"
      },
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "data.head(1).T"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 707
        },
        "id": "FBo8hFlLxBWV",
        "outputId": "da015bcf-e251-4f09-fe43-3db9a3f4dcf9"
      },
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                                      0\n",
              "state                                KS\n",
              "account_length                      128\n",
              "area_code                           415\n",
              "phone_number                   382-4657\n",
              "intl_plan                            no\n",
              "voice_mail_plan                     yes\n",
              "number_vmail_messages                25\n",
              "total_day_minutes                 265.1\n",
              "total_day_calls                     110\n",
              "total_day_charge                  45.07\n",
              "total_eve_minutes                 197.4\n",
              "total_eve_calls                      99\n",
              "total_eve_charge                  16.78\n",
              "total_night_minutes               244.7\n",
              "total_night_calls                    91\n",
              "total_night_charge                11.01\n",
              "total_intl_minutes                 10.0\n",
              "total_intl_calls                      3\n",
              "total_intl_charge                   2.7\n",
              "number_customer_service_calls         1\n",
              "churned                           False"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-264c0e87-389d-49d1-9a61-a82d9ecde91a\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>0</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>state</th>\n",
              "      <td>KS</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>account_length</th>\n",
              "      <td>128</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>area_code</th>\n",
              "      <td>415</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>phone_number</th>\n",
              "      <td>382-4657</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>intl_plan</th>\n",
              "      <td>no</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>voice_mail_plan</th>\n",
              "      <td>yes</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>number_vmail_messages</th>\n",
              "      <td>25</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>total_day_minutes</th>\n",
              "      <td>265.1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>total_day_calls</th>\n",
              "      <td>110</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>total_day_charge</th>\n",
              "      <td>45.07</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>total_eve_minutes</th>\n",
              "      <td>197.4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>total_eve_calls</th>\n",
              "      <td>99</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>total_eve_charge</th>\n",
              "      <td>16.78</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>total_night_minutes</th>\n",
              "      <td>244.7</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>total_night_calls</th>\n",
              "      <td>91</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>total_night_charge</th>\n",
              "      <td>11.01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>total_intl_minutes</th>\n",
              "      <td>10.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>total_intl_calls</th>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>total_intl_charge</th>\n",
              "      <td>2.7</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>number_customer_service_calls</th>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>churned</th>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-264c0e87-389d-49d1-9a61-a82d9ecde91a')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-264c0e87-389d-49d1-9a61-a82d9ecde91a button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-264c0e87-389d-49d1-9a61-a82d9ecde91a');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Remove extraneous columns\n",
        "data.drop(['state', 'area_code', 'phone_number'], axis=1, inplace=True)"
      ],
      "metadata": {
        "id": "zLaeUT9o3ifQ"
      },
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "data.columns"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lIdp5Oo13nL4",
        "outputId": "90ba4c25-39f7-45f1-ee88-78c62025f7c0"
      },
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Index(['account_length', 'intl_plan', 'voice_mail_plan',\n",
              "       'number_vmail_messages', 'total_day_minutes', 'total_day_calls',\n",
              "       'total_day_charge', 'total_eve_minutes', 'total_eve_calls',\n",
              "       'total_eve_charge', 'total_night_minutes', 'total_night_calls',\n",
              "       'total_night_charge', 'total_intl_minutes', 'total_intl_calls',\n",
              "       'total_intl_charge', 'number_customer_service_calls', 'churned'],\n",
              "      dtype='object')"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Question 2\n",
        "* Notice that some of the columns are categorical data and some are floats. These features will need to be numerically encoded using one of the methods from the lecture.\n",
        "* Finally, remember from the lecture that K-nearest neighbors requires scaled data. Scale the data using one of the scaling methods discussed in the lecture."
      ],
      "metadata": {
        "id": "6t3OoFuH4m2e"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "lb = LabelBinarizer()\n",
        "\n",
        "for col in ['intl_plan', 'voice_mail_plan', 'churned']:\n",
        "    data[col] = lb.fit_transform(data[col])"
      ],
      "metadata": {
        "id": "LiGlu8964hdL"
      },
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Mute the sklearn warning\n",
        "import warnings\n",
        "warnings.filterwarnings('ignore', module='sklearn')\n",
        "\n",
        "msc = MinMaxScaler()\n",
        "\n",
        "data = pd.DataFrame(msc.fit_transform(data),  # this is an np.array, not a dataframe.\n",
        "                    columns=data.columns)"
      ],
      "metadata": {
        "id": "8MGlF9cp40x9"
      },
      "execution_count": 10,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Question 3\n",
        "* Separate the feature columns (everything except churned) from the label (churned). This will create two tables.\n",
        "* Fit a K-nearest neighbors model with a value of k=3 to this data and predict the outcome on the same data."
      ],
      "metadata": {
        "id": "YRsHbRn848c-"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Get a list of all the columns that don't contain the label\n",
        "x_cols = [x for x in data.columns if x != 'churned']\n",
        "\n",
        "# Split the data into two dataframes\n",
        "X_data = data[x_cols]\n",
        "y_data = data['churned']\n",
        "\n",
        "# # alternatively:\n",
        "# X_data = data.copy()\n",
        "# y_data = X_data.pop('churned')"
      ],
      "metadata": {
        "id": "peV7Q1nA43tW"
      },
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "knn = KNeighborsClassifier(n_neighbors=3)\n",
        "\n",
        "knn = knn.fit(X_data, y_data)\n",
        "\n",
        "y_pred = knn.predict(X_data)"
      ],
      "metadata": {
        "id": "F1tFo8Pp5E9h"
      },
      "execution_count": 12,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Question 4\n",
        "Ways to measure error haven't been discussed in class yet, but accuracy is an easy one to understand--it is simply the percent of labels that were correctly predicted (either true or false).\n",
        "\n",
        "* Write a function to calculate accuracy using the actual and predicted labels.\n",
        "* Using the function, calculate the accuracy of this K-nearest neighbors model on the data."
      ],
      "metadata": {
        "id": "Z8481iMf5R0N"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Function to calculate the % of values that were correctly predicted\n",
        "\n",
        "def accuracy(real, predict):\n",
        "    return sum(y_data == y_pred) / float(real.shape[0])"
      ],
      "metadata": {
        "id": "ZXD3vWjX5Hwn"
      },
      "execution_count": 13,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(accuracy(y_data, y_pred))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZUpTdA_I5n4x",
        "outputId": "f347246a-1ee8-4197-9ca6-3256fa4f743b"
      },
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.9422\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Question 5\n",
        "* Fit the K-nearest neighbors model again with n_neighbors=3 but this time use distance for the weights. Calculate the accuracy using the function you created above.\n",
        "* Fit another K-nearest neighbors model. This time use uniform weights but set the power parameter for the Minkowski distance metric to be 1 (p=1) i.e. Manhattan Distance.\n",
        "\n",
        "When weighted distances are used for part 1 of this question, a value of 1.0 should be returned for the accuracy. Why do you think this is? Hint: we are predicting on the data and with KNN the model is the data. We will learn how to avoid this pitfall in the next lecture."
      ],
      "metadata": {
        "id": "eRghvrLJ8Tdq"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "knn = KNeighborsClassifier(n_neighbors=5, weights='distance')\n",
        "\n",
        "knn = knn.fit(X_data, y_data)\n",
        "\n",
        "y_pred = knn.predict(X_data)\n",
        "\n",
        "print(accuracy(y_data, y_pred))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "CRtnIzsd8OHT",
        "outputId": "89a19003-ad44-470c-bc76-78d9fc9c25c4"
      },
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1.0\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "knn = KNeighborsClassifier(n_neighbors=5, p=2)\n",
        "\n",
        "knn = knn.fit(X_data, y_data)\n",
        "\n",
        "y_pred = knn.predict(X_data)\n",
        "\n",
        "print(accuracy(y_data, y_pred))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "v8hB5fwu8dV2",
        "outputId": "9d4f4908-370b-4375-bb82-adebd0faff60"
      },
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.9284\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Question 6\n",
        "\n",
        "* Fit a K-nearest neighbors model using values of k (n_neighbors) ranging from 1 to 20. Use uniform weights (the default). The coefficient for the Minkowski distance (p) can be set to either 1 or 2--just be consistent. Store the accuracy and the value of k used from each of these fits in a list or dictionary.\n",
        "* Plot (or view the table of) the accuracy vs k. What do you notice happens when k=1? Why do you think this is? Hint: it's for the same reason discussed above."
      ],
      "metadata": {
        "id": "gRLyEy6Z-hZh"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Fit the K-nearest neighbors model with different values of k\n",
        "# Store the accuracy measurement for each k\n",
        "\n",
        "score_list = list()\n",
        "\n",
        "for k in range(1, 21):\n",
        "    \n",
        "    knn = KNeighborsClassifier(n_neighbors=k)\n",
        "    knn = knn.fit(X_data, y_data)\n",
        "    \n",
        "    y_pred = knn.predict(X_data)\n",
        "    score = accuracy(y_data, y_pred)\n",
        "    \n",
        "    score_list.append((k, score))\n",
        "    \n",
        "score_df = pd.DataFrame(score_list, columns=['k', 'accuracy'])"
      ],
      "metadata": {
        "id": "rOn8feVl8hAY"
      },
      "execution_count": 17,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Import libraries to make the plot\n",
        "\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "\n",
        "%matplotlib inline"
      ],
      "metadata": {
        "id": "Btmz1qQC-0Kq"
      },
      "execution_count": 18,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "sns.set_context('talk')\n",
        "sns.set_style('ticks')\n",
        "sns.set_palette('dark')\n",
        "\n",
        "ax = score_df.set_index('k').plot()\n",
        "\n",
        "ax.set(xlabel='k', ylabel='accuracy')\n",
        "ax.set_xticks(range(1, 21));"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 301
        },
        "id": "Y8NrV747_FdN",
        "outputId": "11d7ef85-8e90-49da-a916-1412c8278766"
      },
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "wS2zJoZ5ANb9"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}