def dtree_classifier(X_train, y_train, X_test, y_test, plotting):
clf = DecisionTreeClassifier(random_state=3169, max_depth=3)
st = time.time()
clf.fit(X_train, y_train)
end = time.time()
train_time = end - st
stt = time.time()
y_pred = clf.predict(X_test)
endt = time.time()
y_test = np.array(y_test)
y_pred = np.array(y_pred)
title = 'Confusion Matrix with Decision Tree'
#plot_tree(clf.fit(X_train, y_train),filled=True)
plt.show()
clf_name = 'dtree'
if plotting == True:
plot_learning_curve(clf,
'Learning Curve for Decision Tree',
X_train,
y_train, (0.7, 1.01),
n_jobs=4)
plot_validation_curve(X_train, y_train, clf, clf_name)
confusion(y_test, y_pred, title)
print('Accuracy for Decision Tree classifier is ' +
str(accuracy_score(y_test, y_pred)))
print('Training time for Decision Tree classifier: ' + str(train_time) +
' seconds')
print('Testing time for Decision Tree classifier: ' + str(endt - stt) +
' seconds')
print()
def knn_classifier(X_train, y_train, X_test, y_test, neighbors, plotting):
clf = KNeighborsClassifier(n_neighbors=2)
st = time.time()
clf.fit(X_train, y_train)
end = time.time()
train_time = end - st
stt = time.time()
y_pred = clf.predict(X_test)
endt = time.time()
y_test = np.array(y_test)
y_pred = np.array(y_pred)
title = 'Confusion Matrix with KNN'
clf_name = 'knn'
if plotting == True:
plot_learning_curve(clf,
'Learning Curve for KNN',
X_train,
y_train, (0.7, 1.01),
n_jobs=4)
plot_validation_curve(X_train, y_train, clf, clf_name)
confusion(y_test, y_pred, title)
print('Accuracy for KNN classifier is ' +
str(accuracy_score(y_test, y_pred)))
print('Training time for KNN classifier: ' + str(train_time) + ' seconds')
print('Testing time for KNN SVM classifier: ' + str(endt - stt) +
' seconds')
print()
def neural_net(X_train, y_train, X_test, y_test, learning_rate, plotting):
clf = MLPClassifier(activation='tanh',
alpha=1e-03,
batch_size='auto',
learning_rate='adaptive',
learning_rate_init=learning_rate,
solver='adam')
st = time.time()
clf.fit(X_train, y_train)
end = time.time()
train_time = end - st
stt = time.time()
y_pred = clf.predict(X_test)
endt = time.time()
y_test = np.array(y_test)
y_pred = np.array(y_pred)
title = 'Confusion Matrix with Neural Network'
clf_name = 'neural_net'
if plotting == True:
plot_learning_curve(clf,
'Learning Curve for Neural Network',
X_train,
y_train, (0.7, 1.01),
n_jobs=4)
plot_validation_curve(X_train, y_train, clf, clf_name)
confusion(y_test, y_pred, title)
print('Accuracy for Neural Network is ' +
str(accuracy_score(y_test, y_pred)))
print('Training time for Neural Network: ' + str(train_time) + ' seconds')
print('Testing time for Neural Network: ' + str(endt - stt) + ' seconds')
print()
def svm_classifier(X_train, y_train, X_test, y_test, kernel, gamma, plotting):
# Apply SVM
clf = svm.SVC(gamma='auto', random_state=3169, kernel=kernel, C=4)
st = time.time()
clf.fit(X_train, y_train)
end = time.time()
train_time = end - st
stt = time.time()
y_pred = clf.predict(X_test)
endt = time.time()
y_test = np.array(y_test)
y_pred = np.array(y_pred)
title = 'Confusion Matrix with SVM'
clf_name = 'svm'
print(str(clf.get_params))
if plotting == True:
plot_learning_curve(clf,
'Learning Curve for SVM',
X_train,
y_train, (0.7, 1.01),
n_jobs=5)
plot_validation_curve(X_train, y_train, clf, clf_name)
confusion(y_test, y_pred, title)
print('Accuracy for SVM classifier is ' +
str(accuracy_score(y_test, y_pred)))
print('Training time for SVM classifier: ' + str(train_time) + ' seconds')
print('Testing time for SVM classifier: ' + str(endt - stt) + ' seconds')
def boost_classifier(X_train, y_train, X_test, y_test, plotting):
clf = DecisionTreeClassifier(random_state=3169,
max_depth=2,
min_samples_leaf=1)
boost_clf = AdaBoostClassifier(base_estimator=clf,
random_state=3169,
n_estimators=100)
st = time.time()
boost_clf.fit(X_train, y_train)
end = time.time()
train_time = end - st
stt = time.time()
y_pred = boost_clf.predict(X_test)
endt = time.time()
y_test = np.array(y_test)
y_pred = np.array(y_pred)
title = 'Confusion Matrix with Boosting'
#plot_tree(clf.fit(X_train, y_train),filled=True)
plt.show()
clf_name = 'boosting'
if plotting == True:
plot_learning_curve(boost_clf,
'Learning Curve for Boosting',
X_train,
y_train, (0.7, 1.01),
n_jobs=4)
plot_validation_curve(X_train, y_train, boost_clf, clf_name)
confusion(y_test, y_pred, title)
print('Accuracy for Boosting classifier is ' +
str(accuracy_score(y_test, y_pred)))
print('Training time for Boosting classifier: ' + str(train_time) +
' seconds')
print('Testing time for Boosting classifier: ' + str(endt - stt) +
' seconds')
print()