def model_01():
"""
猫图像识别，两层网络结构
:return:
"""
X_train, Y_train, X_test, Y_test, classes = load_data() # 猫图像数据
# 模型参数设置
layers_dims = [X_train.shape[0], 7, 1]
num_iter = 2500
learning_rate = 0.0075
print_cost = True
initialization = "sqrt_n"
parameters, costs = basic_model(X_train,
Y_train,
layers_dims=layers_dims,
num_iter=num_iter,
lr=learning_rate,
print_cost=print_cost,
initialization=initialization)
# 预测及评估
prediction_train = predict(parameters, X_train)
prediction_test = predict(parameters, X_test)
print("Train准确率: {}".format(evaluate(prediction_train, Y_train)))
print("test准确率: {}".format(evaluate(prediction_test, Y_test)))
costs_draw(costs, learning_rate=learning_rate)
def model_03():
"""
单隐层平面数据分类
:return:
"""
# 加载数据集
planar = load_planar_dataset()
noisy_circles, noisy_moons, blobs, gaussian_quantiles, no_structure = load_extra_datasets(
)
datasets = {
"planar": planar,
"noisy_circles": noisy_circles,
"noisy_moons": noisy_moons,
"blobs": blobs,
"gaussian_quantiles": gaussian_quantiles
}
data_set = "planar" # 选择数据集
X_train, Y_train = datasets[data_set]
if data_set != "planar":
X_train, Y_train = X_train.T, Y_train.reshape(1, Y_train.shape[0])
if data_set == "blobs":
Y_train = Y_train % 2
# 绘制散点图
data_draw(X_train, Y_train)
# 模型参数设置
layers_dims = [X_train.shape[0], 4, Y_train.shape[0]]
num_iter = 10000
learning_rate = 1.2
print_cost = True
initialization = "random_small"
parameters, costs = basic_model(X_train,
Y_train,
layers_dims=layers_dims,
num_iter=num_iter,
lr=learning_rate,
print_cost=print_cost,
initialization=initialization)
# Plot the decision boundary
plot_decision_boundary(lambda x: predict(parameters, x.T), X_train,
Y_train)
plt.title("Decision Boundary for hidden layer size " + str(4))
plt.show()
# 预测及评估
prediction_train = predict(parameters, X_train)
print("Train准确率: {}".format(evaluate(prediction_train, Y_train)))
costs_draw(costs, learning_rate=learning_rate)
def model_00():
"""
猫图像识别，逻辑回归模型
:return:
"""
# 加载数据
X_train, Y_train, X_test, Y_test, classes = load_data()
# 模型参数设置
layers_dims = [X_train.shape[0], 1]
num_iter = 2000
learning_rate = 0.005
print_cost = True
initialization = "zeros"
# 调用模型
parameters, costs = basic_model(X_train,
Y_train,
layers_dims=layers_dims,
num_iter=num_iter,
lr=learning_rate,
print_cost=print_cost,
initialization=initialization)
costs_draw(costs, learning_rate=learning_rate)
# 预测及评估
prediction_train = predict(parameters, X_train)
prediction_test = predict(parameters, X_test)
print("Train准确率: {}".format(evaluate(prediction_train, Y_train)))
print("test准确率: {}".format(evaluate(prediction_test, Y_test)))
# 预测新图像
new_image = "images/my_image.jpg"
image_reshape = read_image(new_image, show_image=True)
new_image_prediction = predict(parameters, image_reshape)
print("新图像预测值 y = {}".format(int(np.squeeze(new_image_prediction))))
# 不同学习率
different_lr(X_train, Y_train, X_test, Y_test)
def model_reg(lambd, keep_prob, title):
# 加载数据
X_train, Y_train, X_test, Y_test = load_2D_dataset()
# 模型参数设置
layers_dims = [X_train.shape[0], 20, 3, 1]
num_iter = 30000
learning_rate = 0.3
print_cost = False
initialization = "sqrt_n"
# 调用模型
parameters, costs = basic_model(X_train,
Y_train,
layers_dims=layers_dims,
num_iter=num_iter,
lr=learning_rate,
print_cost=print_cost,
initialization=initialization,
lambd=lambd,
keep_prob=keep_prob)
costs_draw(costs, learning_rate=learning_rate)
# 预测及评估
prediction_train = predict(parameters, X_train)
prediction_test = predict(parameters, X_test)
print("Train准确率: {}".format(evaluate(prediction_train, Y_train)))
print("test准确率: {}".format(evaluate(prediction_test, Y_test)))
plt.title(title)
axes = plt.gca()
axes.set_xlim([-0.75, 0.40])
axes.set_ylim([-0.75, 0.65])
plot_decision_boundary(lambda x: predict(parameters, x.T), X_train,
Y_train)
def model_02():
"""猫图像识别，四层网络结构
:return:
"""
X_train, Y_train, X_test, Y_test, classes = load_data() # 猫图像数据
# 模型参数设置
layers_dims = [X_train.shape[0], 20, 7, 5, 1]
num_iter = 2500
learning_rate = 0.0075
print_cost = True
initialization = "sqrt_n"
parameters, costs = basic_model(X_train,
Y_train,
layers_dims=layers_dims,
num_iter=num_iter,
lr=learning_rate,
print_cost=print_cost,
initialization=initialization)
# 预测及评估
prediction_train = predict(parameters, X_train)
prediction_test = predict(parameters, X_test)
print("Train准确率: {}".format(evaluate(prediction_train, Y_train)))
print("test准确率: {}".format(evaluate(prediction_test, Y_test)))
costs_draw(costs, learning_rate=learning_rate)
# 预测新图像
new_image = "images/my_image.jpg"
image_reshape = read_image(new_image, show_image=True)
new_image_prediction = predict(parameters, image_reshape)
print("新图像预测值 y = {}".format(int(np.squeeze(new_image_prediction))))