tenseroflow 拟合 y = ax*x+b
构建神经网络主要分为 4 个步骤:
构造数据、构建网络、训练模型、评估及预测模型。此外,还介绍了一些超参数设定的经验和
技巧
#coding=utf-8
import tensorflow as tf
import numpy as np
x_data = np.linspace(-1,1,300)[:,np.newaxis]
noise = np.random.normal(0,0.05,x_data.shape)
y_data = np.square(x_data) - 0.5 + noise
#神经网络输入变量
xs = tf.placeholder(tf.float32,[None,1])
ys = tf.placeholder(tf.float32,[None,1])
def add_layer(inputs, in_size,out_size,activation_function=None):
#权重
weights = tf.Variable(tf.random_normal([in_size,out_size]))
#偏移
biases = tf.Variable(tf.zeros([1,out_size]) + 0.1)
#
Wx_plus_b = tf.matmul(inputs,weights) + biases
if activation_function is None:
outputs = Wx_plus_b
else :
outputs = activation_function(Wx_plus_b)
return outputs
#hide 20隐藏层
h1 = add_layer(xs,1,20,activation_function = tf.nn.relu)
#output 1输出层 个人感觉关系是 1-》20-》1
prediction = add_layer(h1,20,1,activation_function=None)
#loss
loss = tf.reduce_mean(tf.reduce_sum(tf.square(ys - prediction),reduction_indices=[1]))
train_step = tf.train.GradientDescentOptimizer(0.1).minimize(loss)
#train
init = tf.global_variables_initializer()
sess = tf.Session()
sess.run(init)
for i in range(1000):
sess.run(train_step,feed_dict={xs: x_data,ys:y_data})
if i % 50 == 0:
print(sess.run(loss,feed_dict={xs:x_data,ys:y_data}))