TensorFlow2.0之过拟合问题实战
import tensorflow as tf
from sklearn.datasets import make_moons
from sklearn.model_selection import train_test_split
from matplotlib import pyplot as plt
import seaborn as sns
import matplotlib.markers as mmarkers
import numpy as np
from tensorflow.keras import layers, Sequential, regularizers
from mpl_toolkits.mplot3d import Axes3D
from tensorflow_core.python.keras.layers import Dense
plt.rcParams['font.size'] = 16
plt.rcParams['font.family'] = ['STKaiti']
plt.rcParams['axes.unicode_minus'] = False
OUTPUT_DIR = '../images'
N_EPOCHS = 500
def load_dataset():
# 采样点数
N_SAMPLES = 1000
# 测试数量比率
TEST_SIZE = None
# 从 moon 分布中随机采样 1000 个点,并切分为训练集-测试集
x, y = make_moons(n_samples=N_SAMPLES, noise=0.25, random_state=100)
x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=TEST_SIZE, random_state=42)
return x, y, x_train, x_test, y_train, y_test
def mscatter(x, y, ax=None, m=None, **kw):
if not ax: ax = plt.gca()
sc = ax.scatter(x, y, **kw)
if (m is not None) and (len(m) == len(x)):
paths = []
for marker in m:
if isinstance(marker, mmarkers.MarkerStyle):
marker_obj = marker
else:
marker_obj = mmarkers.MarkerStyle(marker)
path = marker_obj.get_path().transformed(
marker_obj.get_transform())
paths.append(path)
sc.set_paths(paths)
return sc
def make_plot(x, y, plot_name, file_name, xx=None, yy=None, preds=None, dark=False, output_dir=OUTPUT_DIR):
# 绘制数据集的分布,x为2D坐标,y是数据点的标签
if dark:
plt.style.use('dark_background')
else:
sns.set_style("whitegrid")
plt.figure()
axes = plt.gca()
axes.set_xlim([-2, 3])
axes.set_ylim([-1.5, 2])
axes = plt.gca()
axes.set(xlabel="$x_1$", ylabel="$x_2$")
plt.title(plot_name, fontsize=20, fontproperties='SimHei')
plt.subplots_adjust(left=0.2)
plt.subplots_adjust(left=0.8)
if xx is not None and yy is not None and preds is not None:
plt.contourf(xx, yy, preds.reshape(xx.shape), 25, alpha=0.08, cmap=plt.cm.Spectral)
plt.contour(xx, yy, preds.reshape(xx.shape), levels=[.5], cmap="Greys", vmin=0, vmax=.6)
# 绘制散点图,根据标签区分颜色
markers = ['o' if i == 1 else 's' for i in y.ravel()]
mscatter(x[:, 0], x[:, 1], c=y.ravel(), s=20, cmap=plt.cm.Spectral, edgecolors='none', m=markers, ax=axes)
# 保存矢量图
plt.savefig(output_dir + '/' + file_name)
plt.close()
def mscatter(x, y, ax=None, m=None, **kw):
if not ax: ax = plt.gca()
sc = ax.scatter(x, y, **kw)
if (m is not None) and (len(m) == len(x)):
paths = []
for marker in m:
if isinstance(marker, mmarkers.MarkerStyle):
marker_obj = marker
else:
marker_obj = mmarkers.MarkerStyle(marker)
path = marker_obj.get_path().transformed(
marker_obj.get_transform())
paths.append(path)
sc.set_paths(paths)
return sc
def network_layers_influence(x_train, y_train):
for n in range(5):
model = Sequential()
model.add(Dense(8, input_dim=2, activation='relu'))
for _ in range(n):
model.add(Dense(32, activation='sigmoid'))
# 创建最末层
model.add(layers.Dense(1, activation='sigmoid'))
model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])
model.fit(x_train, y_train, epochs=N_EPOCHS, verbose=1)
# 可视化的 x 坐标范围为[-2, 3]
xx = np.arange(-2, 3, 0.01)
# 可视化的 y 坐标范围为[-1.5, 2]
yy = np.arange(-1.5, 2, 0.01)
# 生成 x-y 平面采样网格点,方便可视化
XX, YY = np.meshgrid(xx, yy)
preds = model.predict_classes(np.c_[XX.ravel(), YY.ravel()])
title = "网络层数:{0}".format(2 + n)
file = "网络容量_%i.png" % (2 + n)
make_plot(x_train, y_train, title, file, XX, YY, preds)
def dropout_influence(x_train, y_train):
for n in range(5):
model = Sequential()
model.add(Dense(8, input_dim=2, activation='relu'))
count = 0
for _ in range(5):
model.add(Dense(64, activation='relu'))
if count < n:
count += 1
model.add(layers.Dropout(rate=0.5))
model.add(Dense(1, activation='sigmoid'))
model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])
# 绘制不同 Dropout 层数的决策边界曲线
# 可视化的 x 坐标范围为[-2, 3]
xx = np.arange(-2, 3, 0.01)
# 可视化的 y 坐标范围为[-1.5, 2]
yy = np.arange(-1.5, 2, 0.01)
# 生成 x-y 平面采样网格点,方便可视化
XX, YY = np.meshgrid(xx, yy)
preds = model.predict_classes(np.c_[XX.ravel(), YY.ravel()])
title = "无Dropout层" if n == 0 else "{0}层 Dropout层".format(n)
file = "Dropout_%i.png" % n
make_plot(x_train, y_train, title, file, XX, YY, preds, output_dir=OUTPUT_DIR + '/dropout')
def build_model_with_regularization(_lambda):
# 创建带正则化项的神经网络
model = Sequential()
model.add(layers.Dense(8, input_dim=2, activation='relu')) # 不带正则化项
# 2-4层均是带 L2 正则化项
model.add(layers.Dense(256, activation='relu', kernel_regularizer=regularizers.l2(_lambda)))
model.add(layers.Dense(256, activation='relu', kernel_regularizer=regularizers.l2(_lambda)))
model.add(layers.Dense(256, activation='relu', kernel_regularizer=regularizers.l2(_lambda)))
# 输出层
model.add(layers.Dense(1, activation='sigmoid'))
model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy']) # 模型装配
return model
def plot_weights_matrix(model, layer_index, plot_name, plot_title, file_name, output_dir):
# 绘制权值范围函数
# 提取指定层的权值矩阵
weights = model.layers[layer_index].get_weights()[0]
shape = weights.shape
# 生成和权值矩阵等大小的网格坐标
X = np.array(range(shape[1]))
Y = np.array(range(shape[0]))
X, Y = np.meshgrid(X, Y)
# 绘制3D图
fig = plt.figure()
ax = fig.gca(projection='3d')
ax.xaxis.set_pane_color((1.0, 1.0, 1.0, 0.0))
ax.yaxis.set_pane_color((1.0, 1.0, 1.0, 0.0))
ax.zaxis.set_pane_color((1.0, 1.0, 1.0, 0.0))
plt.title(plot_name, fontsize=20, fontproperties='SimHei')
# 绘制权值矩阵范围
ax.plot_surface(X, Y, weights, cmap=plt.get_cmap('rainbow'), linewidth=0)
# 设置坐标轴名
ax.set_xlabel('网格x坐标', fontsize=16, rotation=0, fontproperties='SimHei')
ax.set_ylabel('网格y坐标', fontsize=16, rotation=0, fontproperties='SimHei')
ax.set_zlabel('权值', fontsize=16, rotation=90, fontproperties='SimHei')
# 保存矩阵范围图
plt.savefig(output_dir + "/" + file_name + ".svg")
plt.close(fig)
def regularizers_influence(X_train, y_train):
for _lambda in [1e-5, 1e-3, 1e-1, 0.12, 0.13]: # 设置不同的正则化系数
# 创建带正则化项的模型
model = build_model_with_regularization(_lambda)
# 模型训练
model.fit(X_train, y_train, epochs=N_EPOCHS, verbose=1)
# 绘制权值范围
layer_index = 2
plot_title = "正则化系数:{}".format(_lambda)
file_name = "正则化网络权值_" + str(_lambda)
# 绘制网络权值范围图
plot_weights_matrix(model, layer_index, plot_title, file_name, output_dir=OUTPUT_DIR + '/regularizers')
# 绘制不同正则化系数的决策边界线
# 可视化的 x 坐标范围为[-2, 3]
xx = np.arange(-2, 3, 0.01)
# 可视化的 y 坐标范围为[-1.5, 2]
yy = np.arange(-1.5, 2, 0.01)
# 生成 x-y 平面采样网格点,方便可视化
XX, YY = np.meshgrid(xx, yy)
preds = model.predict_classes(np.c_[XX.ravel(), YY.ravel()])
title = "正则化系数:{}".format(_lambda)
file = "正则化_%g.svg" % _lambda
make_plot(X_train, y_train, title, file, XX, YY, preds, output_dir=OUTPUT_DIR + '/regularizers')
def main():
x, y, x_train, x_test, y_train, y_test = load_dataset()
# 绘制数据集分布
# make_plot(x, y, None, "月牙形状二分类数据集分布.svg")
# 网络层数的影响
network_layers_influence(x_train, y_train)
# Dropout的影响
# dropout_influence(x_train, y_train)
# # 正则化的影响
# regularizers_influence(x_train, y_train)
if __name__ == '__main__':
main()