def run_Multinomial_Naive_Bayes(clf, alpha, count, fold, use_balanced_set,
use_feature_selection):
training_samples, training_labels, validation_samples, validation_labels = samples_and_labels(
count, fold, use_balanced_set, use_feature_selection, True)
fit_and_predict(clf, alpha, count, fold, training_samples, training_labels,
validation_samples, validation_labels, use_balanced_set,
use_feature_selection)
def run_SVM_Linear(clf, C, count, fold, use_balanced_set,
use_feature_selection):
training_samples, training_labels, validation_samples, validation_labels = samples_and_labels(
count, fold, use_balanced_set, use_feature_selection)
fit_and_predict(clf, C, count, fold, training_samples, training_labels,
validation_samples, validation_labels, use_balanced_set,
use_feature_selection)
def run_Random_Forest(clf, num_tree, max_depth, count, fold, use_balanced_set,
use_feature_selection):
training_samples, training_labels, validation_samples, validation_labels = samples_and_labels(
count, fold, use_balanced_set, use_feature_selection)
fit_and_predict(clf, num_tree, max_depth, count, fold, training_samples,
training_labels, validation_samples, validation_labels,
use_balanced_set, use_feature_selection)
def test_SVM_Linear(tempC, count, use_balanced_set, use_feature_selection):
clf = svm.LinearSVC(C=tempC)
training_samples, training_labels, test_samples, test_labels = samples_and_labels(
count, 0, use_balanced_set, use_feature_selection)
clf.fit(training_samples, training_labels)
result = clf.predict(test_samples)
accuracy, precision, recall, specificity, f_score = calculate_precision_recall(
test_labels, result)
create_result_txt_for_roc_and_pr_plots('svm_linear', clf, test_samples,
test_labels, use_feature_selection)
print 'Accuracy = ' + str(accuracy) + '\nPrecision = ' + str(
precision) + '\nRecall = ' + str(recall) + '\nSpecificity = ' + str(
specificity) + '\nF1 Score = ' + str(f_score) + '\n'
def test_Random_Forest(num_tree, max_depth, count, use_balanced_set,
use_feature_selection):
clf = RandomForestClassifier(n_estimators=num_tree, max_depth=max_depth)
training_samples, training_labels, test_samples, test_labels = samples_and_labels(
count, 0, use_balanced_set, use_feature_selection)
clf.fit(training_samples, training_labels)
result = clf.predict(test_samples)
accuracy, precision, recall, specificity, f_score = calculate_precision_recall(
test_labels, result)
create_result_txt_for_roc_and_pr_plots('random_forest', clf, test_samples,
test_labels, use_feature_selection)
print 'Accuracy = ' + str(accuracy) + '\nPrecision = ' + str(
precision) + '\nRecall = ' + str(recall) + '\nSpecificity = ' + str(
specificity) + '\nF1 Score = ' + str(f_score) + '\n'
def test_Multinomial_Naive_bayes(alpha, count, use_balanced_set,
use_feature_selection):
clf = MultinomialNB(alpha=alpha)
training_samples, training_labels, test_samples, test_labels = samples_and_labels(
count, 0, use_balanced_set, use_feature_selection, True)
clf.fit(training_samples, training_labels)
result = clf.predict(test_samples)
accuracy, precision, recall, specificity, f_score = calculate_precision_recall(
test_labels, result)
if not use_feature_selection:
create_result_txt_for_roc_and_pr_plots('multinomial_naive_bayes', clf,
test_samples, test_labels,
use_feature_selection)
print 'Accuracy = ' + str(accuracy) + '\nPrecision = ' + str(
precision) + '\nRecall = ' + str(recall) + '\nSpecificity = ' + str(
specificity) + '\nF1 Score = ' + str(f_score) + '\n'
