def run_neural_network(X_train, X_test, y_train, y_test):
y_train_f = y_to_float(y_train)
y_test_f = y_to_float(y_test)
classifier = MLPClassifier(alpha=1, max_iter=1000)
classifier.fit(X_train, y_train_f)
y_pred = classifier.predict(X_test)
print("Neural Network")
print(confusion_matrix(y_test_f, y_pred))
print(classification_report(y_test_f, y_pred))
print(accuracy_score(y_test_f, y_pred))
def run_random_forest(X_train, X_test, y_train, y_test):
y_train_f = y_to_float(y_train)
y_test_f = y_to_float(y_test)
classifier = RandomForestClassifier(n_estimators=10, random_state=0, n_jobs=-1)
classifier.fit(X_train, y_train_f)
y_pred = classifier.predict(X_test)
print("Random Forest")
print(confusion_matrix(y_test_f, y_pred))
print(classification_report(y_test_f, y_pred))
print(accuracy_score(y_test_f, y_pred))
'''
def run_bernoulliNB(X_train, X_test, y_train, y_test):
y_train_f = y_to_float(y_train)
y_test_f = y_to_float(y_test)
classifier = BernoulliNB()
classifier.fit(X_train, y_train_f)
y_pred = classifier.predict(X_test)
# Confusion matrix
print("BERNOULLI NAIVE BAYES")
print(confusion_matrix(y_test_f, y_pred))
print(classification_report(y_test_f, y_pred))
print(accuracy_score(y_test_f, y_pred))
def run_ada_boost(X_train, X_test, y_train, y_test):
y_train_f = y_to_float(y_train)
y_test_f = y_to_float(y_test)
classifier = AdaBoostClassifier(DecisionTreeClassifier(max_depth=1),
n_estimators=200)
classifier.fit(X_train, y_train_f)
y_pred = classifier.predict(X_test)
print("Ada Boost")
print(confusion_matrix(y_test_f, y_pred))
print(classification_report(y_test_f, y_pred))
print(accuracy_score(y_test_f, y_pred))
def run_multinomialNB(X_train, X_test, y_train, y_test):
y_train_f = y_to_float(y_train)
y_test_f = y_to_float(y_test)
tf_transformer = TfidfTransformer().fit_transform(X_train)
classifier = MultinomialNB()
classifier.fit(tf_transformer, y_train_f)
y_pred = classifier.predict(X_test)
# Confusion matrix
print("MULTINOMIAL NAIVE BAYES")
print(confusion_matrix(y_test_f, y_pred))
print(classification_report(y_test_f, y_pred))
print(accuracy_score(y_test_f, y_pred))
def run_gaussianNB(X_train, X_test, y_train, y_test):
y_train_f = y_to_float(y_train)
y_test_f = y_to_float(y_test)
# Naive Bayes
classifier = GaussianNB()
classifier.fit(X_train, y_train_f)
# predict class
y_pred = classifier.predict(X_test)
# Confusion matrix
print("GAUSSIAN NAIVE BAYES")
print(confusion_matrix(y_test_f, y_pred))
print(classification_report(y_test_f, y_pred))
print(accuracy_score(y_test_f, y_pred))
def run_svm(X_train, X_test, y_train, y_test):
y_train_f = y_to_float(y_train)
y_test_f = y_to_float(y_test)
classifier = svm.LinearSVC(random_state=0, max_iter=1000)
classifier.fit(X_train, y_train_f)
y_pred = classifier.predict(X_test)
y_score = classifier.decision_function(X_test)
print("SVM")
print(confusion_matrix(y_test_f, y_pred))
print(classification_report(y_test_f, y_pred))
print(accuracy_score(y_test_f, y_pred))
#plot_learning_curve(classifier, "SVM Learning Curve", X_train, y_train_f, ylim=(0.6, 1.01), cv=5, n_jobs=-1)
#plot_roc_curve("SVM ROC Curve", y_test_f, classifier.decision_function(X_test))
#plot_pr_curve("SVM Precision Recall Curve", y_test_f, y_pred, classifier.decision_function(X_test))
def run_linear_regression(X_train, X_test, y_train, y_test):
y_train_f = y_to_float(y_train)
y_test_f = y_to_float(y_test)
regr = LinearRegression(n_jobs=-1)
regr.fit(X_train, y_train_f)
y_pred = regr.predict(X_test)
y_pred = y_pred.round().astype(int)
print("LINEAR REGRESSION")
print(confusion_matrix(y_test_f, y_pred))
print(classification_report(y_test_f, y_pred))
print(accuracy_score(y_test_f, y_pred))
# The coefficients
#print('Coefficients: \n', regr.coef_)
# The mean squared error
print("Mean squared error: %.2f" % mean_squared_error(y_test_f, y_pred))
# Explained variance score: 1 is perfect prediction
print('Variance score: %.2f' % r2_score(y_test_f, y_pred))
'''
def run_decision_tree(X_train, X_test, y_train, y_test):
y_train_f = y_to_float(y_train)
y_test_f = y_to_float(y_test)
classifier = DecisionTreeClassifier()
classifier.fit(X_train, y_train_f)
y_pred = classifier.predict(X_test)
print("Decision Tree")
print(confusion_matrix(y_test_f, y_pred))
print(classification_report(y_test_f, y_pred))
print(accuracy_score(y_test_f, y_pred))
'''
plot_learning_curve(classifier, "Decision Tree Learning Curve", X_train, y_train_f, ylim=(0.6, 1.01), cv=5, n_jobs=-1)
plot_roc_curve("Decision Tree ROC Curve", y_test_f, classifier.predict_proba(X_test)[:, 1])
plot_pr_curve("Decision Tree Precision Recall Curve", y_test_f, y_pred, classifier.predict_proba(X_test)[:, 1])
'''
'''
max_depths = np.linspace(1, 32, 32, endpoint=True)
train_results = []
test_results = []
for max_depth in max_depths:
dt = DecisionTreeClassifier(max_depth=max_depth)
dt.fit(X_train, y_train_f)
train_pred = dt.predict(X_train)
false_positive_rate, true_positive_rate, thresholds = roc_curve(y_train_f, train_pred)
roc_auc = auc(false_positive_rate, true_positive_rate)
# Add auc score to previous train results
train_results.append(roc_auc)
y_pred = dt.predict(X_test)
false_positive_rate, true_positive_rate, thresholds = roc_curve(y_test_f, y_pred)
roc_auc = auc(false_positive_rate, true_positive_rate)
# Add auc score to previous test results
test_results.append(roc_auc)
from matplotlib.legend_handler import HandlerLine2D
line1, = plt.plot(max_depths, train_results, 'b', label="Train AUC")
line2, = plt.plot(max_depths, test_results, 'r', label="Test AUC")
plt.legend(handler_map={line1: HandlerLine2D(numpoints=2)})
plt.ylabel('AUC score')
plt.xlabel('Tree depth')
plt.show()
'''
'''
min_samples_splits = np.linspace(0.1, 1.0, 10, endpoint=True)
train_results = []
test_results = []
for min_samples_split in min_samples_splits:
dt = DecisionTreeClassifier(min_samples_split=min_samples_split)
dt.fit(X_train, y_train_f)
train_pred = dt.predict(X_train)
false_positive_rate, true_positive_rate, thresholds = roc_curve(y_train_f, train_pred)
roc_auc = auc(false_positive_rate, true_positive_rate)
train_results.append(roc_auc)
y_pred = dt.predict(X_test)
false_positive_rate, true_positive_rate, thresholds = roc_curve(y_test_f, y_pred)
roc_auc = auc(false_positive_rate, true_positive_rate)
test_results.append(roc_auc)
from matplotlib.legend_handler import HandlerLine2D
line1, = plt.plot(min_samples_splits, train_results, 'b', label="Train AUC")
line2, = plt.plot(min_samples_splits, test_results, 'r', label="Test AUC")
plt.legend(handler_map={line1: HandlerLine2D(numpoints=2)})
plt.ylabel('AUC score')
plt.xlabel('min samples split')
plt.show()
'''
'''