def main_adaboost():
"""
Main function for testing Adaboost.
"""
X_train, Y_train, X_test, Y_test = h_shaped_dataset()
# X_train, Y_train, X_test, Y_test = gaussians_dataset(2, [100, 150], [[1, 3], [-4, 8]], [[2, 3], [4, 1]])
# X_train, Y_train, X_test, Y_test = two_moon_dataset(n_samples=300, noise=0.2)
# visualize dataset
plot_2d_dataset(X_train, Y_train, 'Training')
# train model and predict
model = AdaBoostClassifier(n_learners=100)
model.fit(X_train, Y_train, verbose=True)
P = model.predict(X_test)
# visualize the boundary!
plot_boundary(X_train, Y_train, model)
# evaluate and print error
error = float(np.sum(P != Y_test)) / Y_test.size
print('Error: {}, Accuracy (1-Error): {}'.format(error, 1 - error))
def main_adaboost():
"""
Main function for fitting and testing Adaboost classifier.
"""
X_train, Y_train, X_test, Y_test = gaussians_dataset(2, [300, 400], [[1, 3], [-4, 8]], [[2, 3], [4, 1]])
# X_train, Y_train, X_test, Y_test = h_shaped_dataset()
# X_train, Y_train, X_test, Y_test = two_moon_dataset(n_samples=500, noise=0.2)
# Visualize dataset
plot_2d_dataset(X_train, Y_train, 'Training', blocking=False)
# Init model
model = AdaBoostClassifier(n_learners=100)
# Train
model.fit(X_train, Y_train, verbose=True)
# Predict
y_preds = model.predict(X_test)
print('Accuracy on test set: {}'.format(np.mean(y_preds == Y_test)))
# Visualize the predicted boundary
plot_boundary(X_train, Y_train, model)
def main():
"""
分别训练不使用正则化的原始模型、使用L2正则化的模型、使用dropout的模型，输出cost值、预测准确率、边界预测结果等信息
"""
# 读入数据集
train_X, train_Y, test_X, test_Y = utils.load_dataset("data")
# 训练不使用正则化的原始模型
model_name = "_without_regularization"
parameters = model(train_X, train_Y, model_name)
# 在训练集和测试集上的进行预测并输出准确率
print "On the training set:"
predictions_train = utils.predict(train_X, train_Y, parameters)
print "On the test set:"
predictions_test = utils.predict(test_X, test_Y, parameters)
# 绘制边界预测结果图
plt.clf()
plt.title("Model without regularization")
axes = plt.gca()
axes.set_xlim([-0.75, 0.40])
axes.set_ylim([-0.75, 0.65])
utils.plot_boundary(lambda x: utils.predict_decision(parameters, x.T),
train_X, train_Y,
"decision_boundary" + model_name + ".png")
# 训练使用L2正则化的模型
model_name = "_with_L2-regularization"
parameters = model(train_X, train_Y, model_name, lambd=0.7)
# 在训练集和测试集上的进行预测并输出准确率
print "On the train set:"
predictions_train = utils.predict(train_X, train_Y, parameters)
print "On the test set:"
predictions_test = utils.predict(test_X, test_Y, parameters)
# 绘制边界预测结果图
plt.clf()
plt.title("Model with L2-regularization")
axes = plt.gca()
axes.set_xlim([-0.75, 0.40])
axes.set_ylim([-0.75, 0.65])
utils.plot_boundary(lambda x: utils.predict_decision(parameters, x.T),
train_X, train_Y,
"decision_boundary" + model_name + ".png")
# 训练使用dropout的模型
model_name = "_with_dropout"
parameters = model(train_X, train_Y, model_name, keep_prob=0.86)
# 在训练集和测试集上的进行预测并输出准确率
print "On the train set:"
predictions_train = utils.predict(train_X, train_Y, parameters)
print "On the test set:"
predictions_test = utils.predict(test_X, test_Y, parameters)
# 绘制边界预测结果图
plt.clf()
plt.title("Model with dropout")
axes = plt.gca()
axes.set_xlim([-0.75, 0.40])
axes.set_ylim([-0.75, 0.65])
utils.plot_boundary(lambda x: utils.predict_decision(parameters, x.T),
train_X, train_Y,
"decision_boundary" + model_name + ".png")