LightGBM working

README.md

@@ -1,17 +1,18 @@
 ## Overview
 
-This project is my implementation of [this research paper](https://ernest-bonat.medium.com/rna-seq-gene-expression-classification-using-machine-learning-algorithms-de862e60bfd0#4592) from scratch. The objective of this project is to identify gene expression patterns associated with different conditions or diseases, leveraging advanced data processing and model training techniques. RNA sequencing (RNA-Seq) is a teachnique used to quantify and analyze gene expression levels across different conditions or samples. The classification of RNA-Seq data can be used to identify which genes are differently expressed between healthy and diseased samples, or between different diseased states, thus aiding in the diagnosis, treatment, and understanding of various diseases and conditions.
+The objective of this project is to identify gene expression patterns associated with different conditions or diseases, leveraging advanced data processing and model training techniques. RNA sequencing (RNA-Seq) is a teachnique used to quantify and analyze gene expression levels across different conditions or samples. The classification of RNA-Seq data can be used to identify which genes are differently expressed between healthy and diseased samples, or between different diseased states, thus aiding in the diagnosis, treatment, and understanding of various diseases and conditions.
 
 ## Table of Contents
 
 - [Overview](#overview)
+- [Table of Contents](#table-of-contents)
 - [Dataset](#dataset)
 - [Project Structure](#project-structure)
-- [Running the project](#running-the-project)
+- [Running the Project](#running-the-project)
 - [Models Used](#models-used)
 - [Results](#results)
 - [Contributing](#contributing)
-- [Acknowledgements](#acknowledgements)
+- [**References**](#references)
 
 ## Dataset
 
@@ -163,5 +164,5 @@ If you find any issues or have suggestions for improvements or expanding the pro
 4. Push to the branch (`git push origin feature-branch`).
 5. Create a new Pull Request.
 
-## **Acknowledgements**
-
+## **References**
+1. [RNA-Seq Gene Expression Classification Using Machine Learning Algorithms](https://ernest-bonat.medium.com/rna-seq-gene-expression-classification-using-machine-learning-algorithms-de862e60bfd0) 

results/svm.txt

@@ -0,0 +1,41 @@
+Validation Accuracy: 0.98
+Validation Classification Report:
+              precision    recall  f1-score   support
+
+        BRCA       0.97      1.00      0.98        28
+        COAD       1.00      1.00      1.00         9
+        KIRC       1.00      0.90      0.95        10
+        LUAD       0.92      0.92      0.92        13
+        PRAD       1.00      1.00      1.00        21
+
+    accuracy                           0.98        81
+   macro avg       0.98      0.96      0.97        81
+weighted avg       0.98      0.98      0.98        81
+
+Validation Confusion Matrix:
+[[28  0  0  0  0]
+ [ 0  9  0  0  0]
+ [ 0  0  9  1  0]
+ [ 1  0  0 12  0]
+ [ 0  0  0  0 21]]
+
+Test Accuracy: 0.99
+Test Classification Report:
+              precision    recall  f1-score   support
+
+        BRCA       0.96      1.00      0.98        27
+        COAD       1.00      1.00      1.00         8
+        KIRC       1.00      1.00      1.00        15
+        LUAD       1.00      0.95      0.97        19
+        PRAD       1.00      1.00      1.00        11
+
+    accuracy                           0.99        80
+   macro avg       0.99      0.99      0.99        80
+weighted avg       0.99      0.99      0.99        80
+
+Test Confusion Matrix:
+[[27  0  0  0  0]
+ [ 0  8  0  0  0]
+ [ 0  0 15  0  0]
+ [ 1  0  0 18  0]
+ [ 0  0  0  0 11]]

src/models/light_gbm.py

@@ -24,10 +24,10 @@ def pickle_serialize_object(filename, obj):
 def main():
     # Deserialize the input
     input_dir = 'data/processed'
-    X_train_pca = pickle_deserialize_object(os.path.join(input_dir, 'X_train_pca.pkl'))
     y_train_resampled = pickle_deserialize_object(os.path.join(input_dir, 'y_train_resampled.pkl'))    
 
     input_dir2 = 'data/processed/transformed'
+    X_train_pca = pickle_deserialize_object(os.path.join(input_dir2, 'X_train_pca.pkl'))
     X_val_pca = pickle_deserialize_object(os.path.join(input_dir2, 'X_val_pca.pkl'))
     X_test_pca = pickle_deserialize_object(os.path.join(input_dir2, 'X_test_pca.pkl'))
 
@@ -37,12 +37,16 @@ def main():
 
     # Define parameter grid for GridSearchCV
     param_grid = {
-        'n_estimators': [50, 100],
-        'learning_rate': [0.01],
         'num_leaves': [31],
-        'max_depth': [10, 20],
-        'min_child_samples': [100, 200],
-        'force_col_wise': [True]
+        'min_child_samples': [10],
+        'min_split_gain': [0.01],
+        'learning_rate': [0.1,],
+        'max_depth': [-1, 10, 20],
+        'reg_alpha': [0, 0.1, 0.5],
+        'reg_lambda': [0, 0.1, 0.5],
+        'subsample': [0.8],
+        'colsample_bytree': [0.8],
+        'n_estimators': [100]
 }
 
     # Initialize and fit LGBMClassifier with GridSearchCV
@@ -60,25 +64,26 @@ def main():
     val_confusion_matrix = confusion_matrix(y_val, y_val_pred)
 
     # Evaluate on test set
+    '''
     y_test_pred = best_lgbm.predict(X_test_pca)
     test_accuracy = accuracy_score(y_test, y_test_pred)
     test_classification_report = classification_report(y_test, y_test_pred)
     test_confusion_matrix = confusion_matrix(y_test, y_test_pred)
-
+    '''
     # Write results to a file
-    output_filename = 'results/lightgbm.txt'
+    output_filename = 'results/light_gbm.out'
     with open(output_filename, 'w') as f:
         f.write(f"Validation Accuracy: {val_accuracy:.2f}\n")
         f.write("Validation Classification Report:\n")
         f.write(val_classification_report + '\n')
         f.write("Validation Confusion Matrix:\n")
         f.write(str(val_confusion_matrix) + '\n\n')
-
+    '''    
         f.write(f"Test Accuracy: {test_accuracy:.2f}\n")
         f.write("Test Classification Report:\n")
         f.write(test_classification_report + '\n')
         f.write("Test Confusion Matrix:\n")
         f.write(str(test_confusion_matrix) + '\n')
-
+    '''
 if __name__ == "__main__":
     main()

src/models/random_forest.py

@@ -24,10 +24,10 @@ def pickle_serialize_object(filename, obj):
 def main():
     # Deserialize the input
     input_dir = 'data/processed'
-    X_train_pca = pickle_deserialize_object(os.path.join(input_dir, 'X_train_pca.pkl'))
     y_train_resampled = pickle_deserialize_object(os.path.join(input_dir, 'y_train_resampled.pkl'))    
 
     input_dir2 = 'data/processed/transformed'
+    X_train_pca = pickle_deserialize_object(os.path.join(input_dir2, 'X_train_pca.pkl'))
     X_val_pca = pickle_deserialize_object(os.path.join(input_dir2, 'X_val_pca.pkl'))
     X_test_pca = pickle_deserialize_object(os.path.join(input_dir2, 'X_test_pca.pkl'))
 

src/models/svm_classifier.py

@@ -0,0 +1,82 @@
+import os
+import sys
+import datetime
+import pickle
+from sklearn.svm import SVC
+from sklearn.metrics import accuracy_score, classification_report, confusion_matrix
+from sklearn.model_selection import GridSearchCV
+
+# Generate a timestamp for this run
+timestamp = datetime.datetime.now().strftime("%Y-%m-%d_%H-%M-%S")
+log_file = f"logs/svm_{timestamp}.out"
+
+# Redirect stdout and stderr to the log file
+sys.stdout = open(f'{log_file}', 'a')
+sys.stderr = open(f'{log_file}', 'a')
+
+def pickle_deserialize_object(filename):
+    with open(filename, 'rb') as f:
+        return pickle.load(f)
+
+def pickle_serialize_object(filename, obj):
+    with open(filename, 'wb') as f:
+        pickle.dump(obj, f)
+
+def main():
+    # Deserialize the input
+    input_dir = 'data/processed'
+    X_train_pca = pickle_deserialize_object(os.path.join(input_dir, 'X_train_pca.pkl'))
+    y_train_resampled = pickle_deserialize_object(os.path.join(input_dir, 'y_train_resampled.pkl'))
+
+    input_dir2 = 'data/processed/transformed'
+    X_val_pca = pickle_deserialize_object(os.path.join(input_dir2, 'X_val_pca.pkl'))
+    X_test_pca = pickle_deserialize_object(os.path.join(input_dir2, 'X_test_pca.pkl'))
+
+    input_dir3 = 'data/processed/split_data'
+    y_val = pickle_deserialize_object(os.path.join(input_dir3, 'y_val.pkl'))
+    y_test = pickle_deserialize_object(os.path.join(input_dir3, 'y_test.pkl'))
+
+    # Define parameter grid for GridSearchCV
+    param_grid = {
+        'C': [0.1, 1, 10],
+        'gamma': [0.001, 0.01, 0.1],
+        'kernel': ['rbf', 'linear']
+    }
+
+    # Initialize and fit SVM with GridSearchCV
+    svm = SVC(random_state=1)
+    grid_search = GridSearchCV(svm, param_grid, cv=3, n_jobs=-1, verbose=1)
+    grid_search.fit(X_train_pca, y_train_resampled)
+
+    # Get the best estimator
+    best_svm = grid_search.best_estimator_
+
+    # Evaluate on validation set
+    y_val_pred = best_svm.predict(X_val_pca)
+    val_accuracy = accuracy_score(y_val, y_val_pred)
+    val_classification_report = classification_report(y_val, y_val_pred)
+    val_confusion_matrix = confusion_matrix(y_val, y_val_pred)
+
+    # Evaluate on test set
+    y_test_pred = best_svm.predict(X_test_pca)
+    test_accuracy = accuracy_score(y_test, y_test_pred)
+    test_classification_report = classification_report(y_test, y_test_pred)
+    test_confusion_matrix = confusion_matrix(y_test, y_test_pred)
+
+    # Write results to a file
+    output_filename = 'results/svm.txt'
+    with open(output_filename, 'w') as f:
+        f.write(f"Validation Accuracy: {val_accuracy:.2f}\n")
+        f.write("Validation Classification Report:\n")
+        f.write(val_classification_report + '\n')
+        f.write("Validation Confusion Matrix:\n")
+        f.write(str(val_confusion_matrix) + '\n\n')
+
+        f.write(f"Test Accuracy: {test_accuracy:.2f}\n")
+        f.write("Test Classification Report:\n")
+        f.write(test_classification_report + '\n')
+        f.write("Test Confusion Matrix:\n")
+        f.write(str(test_confusion_matrix) + '\n')
+
+if __name__ == "__main__":
+    main()