add fine tume for SVM

classification.py

@@ -200,6 +200,7 @@ if __name__ == '__main__':
         # Train the classifier using the training data.
         # Test the classifier using the testing data.
         from sklearn.svm import SVC
+        from sklearn.model_selection import GridSearchCV
         from sklearn.metrics import accuracy_score
         from sklearn.metrics import precision_score
         import pandas as pd
@@ -219,18 +220,57 @@ if __name__ == '__main__':
         X_test_lbp = test_lbp.drop('class', axis=1)
         y_test_lbp = test_lbp['class']
 
+        # Define the parameter grid for tuning
+        param_grid = {
+            'C': [0.1, 1, 10, 100],         # Regularization parameter
+            'kernel': ['linear', 'rbf'],     # Kernels to explore: linear and RBF
+            'gamma': [1, 0.1, 0.01, 0.001]  # Gamma values for RBF kernel
+        }
+
         clf_glcm = SVC()
+
+        # Set up GridSearchCV with 5-fold cross-validation
+        grid_search = GridSearchCV(clf_glcm, param_grid, cv=5, scoring='accuracy', verbose=2, n_jobs=-1)
+
+        # Perform the grid search to find the best hyperparameters
+        grid_search.fit(X_train_glcm, y_train_glcm)
+
+        # Output the best parameters from the search
+        print("Best parameters for clf_glcm: ", grid_search.best_params_)
+
+        # Use the best estimator found by GridSearchCV to make predictions
+        clf_glcm = grid_search.best_estimator_
+
         clf_glcm.fit(X_train_glcm, y_train_glcm)
         y_pred_glcm = clf_glcm.predict(X_test_glcm)
+
+        # calculate the accuracy
         print('Accuracy for GLCM features:', accuracy_score(y_test_glcm, y_pred_glcm))
+        
         # calculate the precsion
         precision = precision_score(y_test_glcm, y_pred_glcm, average='weighted')
         print('Precision for GLCM features:', precision)
 
         clf_lbp = SVC()
+        # Set up GridSearchCV with 5-fold cross-validation
+        grid_search = GridSearchCV(clf_lbp, param_grid, cv=5, scoring='accuracy', verbose=2, n_jobs=-1)
+
+        # Perform the grid search to find the best hyperparameters
+        grid_search.fit(X_train_lbp, y_train_lbp)
+
+        # Output the best parameters from the search
+        print("Best parameters for clf_lbp: ", grid_search.best_params_)
+
+        # Use the best estimator found by GridSearchCV to make predictions
+        clf_lbp = grid_search.best_estimator_
+
         clf_lbp.fit(X_train_lbp, y_train_lbp)
+
         y_pred_lbp = clf_lbp.predict(X_test_lbp)
+
+        # calculate the accuracy
         print('Accuracy for LBP features:', accuracy_score(y_test_lbp, y_pred_lbp))
+        
         # calculate the precsion
         precision = precision_score(y_test_lbp, y_pred_lbp, average='weighted')
         print('Precision for LBP features:', precision)
@@ -264,7 +304,7 @@ if __name__ == '__main__':
         plt.savefig('train_lbp_distribution.png')
         plt.close()
 
-        #Use plots to visualize feature distributions and decision boundaries of the classifiers clf_glcm, clf_lbp using t-sne
+        # Use plots to visualize feature distributions and decision boundaries of the classifiers clf_glcm, clf_lbp using t-sne
         from sklearn.manifold import TSNE
 
         tsne = TSNE(n_components=2)
@@ -285,4 +325,23 @@ if __name__ == '__main__':
         plt.title('LBP features')
         # save the plot to a file
         plt.savefig('train_lbp_tsne.png')
+        plt.close()
+
+        # plot t-sne it for the testing data
+        tsne = TSNE(n_components=2, perplexity=5)
+        X_test_glcm_tsne = tsne.fit_transform(X_test_glcm)
+        X_test_lbp_tsne = tsne.fit_transform(X_test_lbp)
+
+        plt.scatter(X_test_glcm_tsne[y_test_glcm == 'grass', 0], X_test_glcm_tsne[y_test_glcm == 'grass', 1], color='red', label='grass')
+        plt.scatter(X_test_glcm_tsne[y_test_glcm == 'wood', 0], X_test_glcm_tsne[y_test_glcm == 'wood', 1], color='blue', label='wood')
+        plt.legend()
+        plt.title('GLCM features')
+        plt.savefig('test_glcm_tsne.png')
+        plt.close()
+
+        plt.scatter(X_test_lbp_tsne[y_test_lbp == 'grass', 0], X_test_lbp_tsne[y_test_lbp == 'grass', 1], color='red', label='grass')
+        plt.scatter(X_test_lbp_tsne[y_test_lbp == 'wood', 0], X_test_lbp_tsne[y_test_lbp == 'wood', 1], color='blue', label='wood')
+        plt.legend()
+        plt.title('LBP features')
+        plt.savefig('test_lbp_tsne.png')
         plt.close()

clf_glcm.pkl

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dae47a92e34746a8ada666bda3481e425b71b3cfc38fadb9bebdbb736bd7e8f5
-size 1838
+oid sha256:87bdb02aa47718d5e9f6ad188b4f02609f826f0b088b5dfaf3ccdbe76551b7ba
+size 1537

clf_lbp.pkl

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2cf77958648f5267635cdda0bb7b5d82b84443a7e9abc8b8dd421faa96d8b0e7
-size 3250
+oid sha256:67242af0119c0f138b1f3d104bbd6144e02e7fe137fa46937714385984d8f14a
+size 2509