本笔记主要记录常见的三个激活函数sigmoid，tanh和relu，关于激活函数详细的描述，可以参考这里：

        详解激活函数（Sigmoid/Tanh/ReLU/Leaky ReLu等） - 知乎

import tensorflow as tf
import numpy as nptf.__version__#详细的激活函数参考资料
#https://zhuanlan.zhihu.com/p/427541517
#Sigmoid/Logistic
#f(x) = 1/(1 + e^(-x))
data = tf.linspace(-10., 10.,  10)
print(data)with tf.GradientTape() as tape:tape.watch(data)y = tf.sigmoid(data)print("y = tf.sigmoid(data) = ", y.numpy())
grads = tape.gradient(y, [data])
print("Sigmoid Gradient:", grads[0].numpy())#Tanh
#f(x) = (e^x - e^-x) / (e^x + e^-x)
data = tf.linspace(-5., 5., 10)
print("Data=", data.numpy())
y = tf.tanh(data)
print("y = tf.tanh(data) = ", y.numpy())#Relu, Rectified Linear Unit
#f(x) = { 0  for x < 0
#         x for x >= 0
data = tf.linspace(-1., 1., 10)
print("Data=", data.numpy())
y = tf.nn.relu(data)
print("y = tf.relu(data) = ", y.numpy())#leaky relu, x < 0时，会得到一个接近0的负值
y_leaky = tf.nn.leaky_relu(data)
print("y_leaky = tf.leaky_relu(data) = ", y_leaky.numpy())

运行结果：