本文将介绍如下内容：
- 自定义损失函数
- 自定义DenseLayer
一，自定义损失函数
1，在TF中实现自定义损失函数
在TF中实现的demo如下：
import tensorflow as tf
import numpy as np
y_pred = np.array([1,4,3,2,6,5,9])
y_true = np.array([1,4,3,2,1,4,9])
print(tf.reduce_mean(tf.square(y_pred - y_true)))
# ---output------
tf.Tensor(3, shape=(), dtype=int64)
结合之前的实例如下：
def customized_mse(y_true, y_pred):
return tf.reduce_mean(tf.square(y_pred - y_true))
model = keras.models.Sequential([
keras.layers.Dense(30, activation='relu',
input_shape=x_train.shape[1:]),
keras.layers.Dense(1),
])
model.summary()
model.compile(loss=customized_mse, optimizer="sgd",
metrics=["mean_squared_error"])
callbacks = [keras.callbacks.EarlyStopping(
patience=5, min_delta=1e-2)]
2，总结代码如下：
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
import matplotlib as mpl
import matplotlib.pyplot as plt
import numpy as np
import sklearn
import pandas as pd
import os
import sys
import time
import tensorflow as tf
from tensorflow import keras
# 1,打印使用的python库的版本信息
print(tf.__version__)
print(sys.version_info)
for module in mpl, np, pd, sklearn, tf, keras:
print(module.__name__, module.__version__)
# 2,下载并使用sklearn中的“fetch_california_housing”数据集
from sklearn.datasets import fetch_california_housing
housing = fetch_california_housing()
print(housing.DESCR)
print(housing.data.shape)
print(housing.target.shape)
# 3,拆分数据集中的数据为 训练数据、验证数据、测试数据
from sklearn.model_selection import train_test_split
x_train_all, x_test, y_train_all, y_test = train_test_split(
housing.data, housing.target, random_state = 7)
x_train, x_valid, y_train, y_valid = train_test_split(
x_train_all, y_train_all, random_state = 11)
print(x_train.shape, y_train.shape)
print(x_valid.shape, y_valid.shape)
print(x_test.shape, y_test.shape)
# 4,在将数据带入到模型之前，先进行预处理-训练、验证、测试数据标准化
from sklearn.preprocessing import StandardScaler
scaler = StandardScaler()
x_train_scaled = scaler.fit_transform(x_train)
x_valid_scaled = scaler.transform(x_valid)
x_test_scaled = scaler.transform(x_test)
# 5,构建模型、模型层级图、编译模型（添加损失函数、优化器）、添加回调函数
def customized_mse(y_true, y_pred):
return tf.reduce_mean(tf.square(y_pred - y_true))
model = keras.models.Sequential([
keras.layers.Dense(30, activation='relu',
input_shape=x_train.shape[1:]),
keras.layers.Dense(1),
])
model.summary()
model.compile(loss=customized_mse, optimizer="sgd",
metrics=["mean_squared_error"])
callbacks = [keras.callbacks.EarlyStopping(
patience=5, min_delta=1e-2)]
# 6,训练构建的模型
history = model.fit(x_train_scaled, y_train,
validation_data = (x_valid_scaled, y_valid),
epochs = 100,
callbacks = callbacks,
verbose=0)
# 7,得到训练曲线图
def plot_learning_curves(history):
pd.DataFrame(history.history).plot(figsize=(8, 5))
plt.grid(True)
plt.gca().set_ylim(0, 1)
plt.show()
plot_learning_curves(history)
# 8,调用估计器-估计测试数据集
print(model.evaluate(x_test_scaled, y_test, verbose=0))
#---output------
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
dense_591 (Dense) (None, 30) 270
_________________________________________________________________
dense_592 (Dense) (None, 1) 31
=================================================================
Total params: 301
Trainable params: 301
Non-trainable params: 0
_________________________________________________________________
[0.3701427737871806, 0.3701428]
二，自定义DenseLayer
1，对于tf.keras.layers.Dense的理解
layer = tf.keras.layers.Dense(100)
layer = tf.keras.layers.Dense(100, input_shape=(None, 5))
temp = layer(tf.zeros([10, 5]))
print(temp.shape)
# ---output-----
(10, 100)
a）这里介绍tf.keras.layers.Dense的两个属性：
tf.keras.layers中的variables和trainable_variables
variables和trainable_variables的数值如下：
print(layer.variables)
print(layer.trainable_variables)
b）tf.keras.layers的计算公式如下:
x * w + b
w 等同于"kernel";b等同于"bias"
c）如果想了解其他的参数详情,可以参看如下：
help(layer)
2，在TF中实现自定义DenseLayer
# 5,自定义激活函数
# tf.nn.softplus : log(1+e^x)
customized_softplus = keras.layers.Lambda(lambda x : tf.nn.softplus(x))
# 输入测试数据测试自定义的激活函数
print(customized_softplus([-10., -5., 0., 5., 10.]))
# 6,自定义DenseLayer函数
# customized dense layer.
class CustomizedDenseLayer(keras.layers.Layer):
def __init__(self, units, activation=None, **kwargs):
self.units = units
self.activation = keras.layers.Activation(activation)
super(CustomizedDenseLayer, self).__init__(**kwargs)
def build(self, input_shape):
"""构建所需要的参数"""
# x * w + b. input_shape:[None, a] w:[a,b]output_shape: [None, b]
self.kernel = self.add_weight(name = 'kernel',
shape = (input_shape[1], self.units),
initializer = 'uniform',
trainable = True)
self.bias = self.add_weight(name = 'bias',
shape = (self.units, ),
initializer = 'zeros',
trainable = True)
super(CustomizedDenseLayer, self).build(input_shape)
def call(self, x):
"""完成正向计算"""
return self.activation(x @ self.kernel + self.bias)
3，总结代码如下：
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
import matplotlib as mpl
import matplotlib.pyplot as plt
import numpy as np
import sklearn
import pandas as pd
import os
import sys
import time
import tensorflow as tf
from tensorflow import keras
# 1,打印使用的python库的版本信息
print(tf.__version__)
print(sys.version_info)
for module in mpl, np, pd, sklearn, tf, keras:
print(module.__name__, module.__version__)
# 2,下载并使用sklearn中的“fetch_california_housing”数据集
# tf.keras.layers传入,输出的参数
from sklearn.datasets import fetch_california_housing
housing = fetch_california_housing()
print(housing.DESCR)
print(housing.data.shape)
print(housing.target.shape)
# 3,拆分数据集中的数据为 训练数据、验证数据、测试数据
from sklearn.model_selection import train_test_split
x_train_all, x_test, y_train_all, y_test = train_test_split(
housing.data, housing.target, random_state = 7)
x_train, x_valid, y_train, y_valid = train_test_split(
x_train_all, y_train_all, random_state = 11)
print(x_train.shape, y_train.shape)
print(x_valid.shape, y_valid.shape)
print(x_test.shape, y_test.shape)
# 4,在将数据带入到模型之前，先进行预处理-训练、验证、测试数据标准化
from sklearn.preprocessing import StandardScaler
scaler = StandardScaler()
x_train_scaled = scaler.fit_transform(x_train)
x_valid_scaled = scaler.transform(x_valid)
x_test_scaled = scaler.transform(x_test)
# 5,自定义激活函数
# tf.nn.softplus : log(1+e^x)
customized_softplus = keras.layers.Lambda(lambda x : tf.nn.softplus(x))
print(customized_softplus([-10., -5., 0., 5., 10.]))
# 6,自定义DenseLayer函数
# customized dense layer.
class CustomizedDenseLayer(keras.layers.Layer):
def __init__(self, units, activation=None, **kwargs):
self.units = units
self.activation = keras.layers.Activation(activation)
super(CustomizedDenseLayer, self).__init__(**kwargs)
def build(self, input_shape):
"""构建所需要的参数"""
# x * w + b. input_shape:[None, a] w:[a,b]output_shape: [None, b]
self.kernel = self.add_weight(name = 'kernel',
shape = (input_shape[1], self.units),
initializer = 'uniform',
trainable = True)
self.bias = self.add_weight(name = 'bias',
shape = (self.units, ),
initializer = 'zeros',
trainable = True)
super(CustomizedDenseLayer, self).build(input_shape)
def call(self, x):
"""完成正向计算"""
return self.activation(x @ self.kernel + self.bias)
# 7,构建模型、自定义模型层级、自定义激活函数、编译模型（添加损失函数、优化器）、添加回调函数
model = keras.models.Sequential([
CustomizedDenseLayer(30, activation='relu',
input_shape=x_train.shape[1:]),
CustomizedDenseLayer(1),
customized_softplus,
# keras.layers.Dense(1, activation="softplus"),
# keras.layers.Dense(1), keras.layers.Activation('softplus'),
])
model.summary()
model.compile(loss="mean_squared_error", optimizer="sgd")
callbacks = [keras.callbacks.EarlyStopping(
patience=5, min_delta=1e-2)]
# 8,训练构建的模型
history = model.fit(x_train_scaled, y_train,
validation_data = (x_valid_scaled, y_valid),
epochs = 100,
callbacks = callbacks)
# 9,得到训练曲线图
def plot_learning_curves(history):
pd.DataFrame(history.history).plot(figsize=(8, 5))
plt.grid(True)
plt.gca().set_ylim(0, 1)
plt.show()
plot_learning_curves(history)
# 10,调用估计器-估计测试数据集
model.evaluate(x_test_scaled, y_test, verbose=0)
# ---output------
0.38935184751370155