Tensorflow学习：单层神经网络的建立

本文内容：简单的建立一个神经网络
注：由于前博客中已经写过了线性回归以及深度神经网络（cnn）的博客，这篇看起来更简单些。

# -*- coding: utf-8 -*-
"""
Created on Wed May  3 11:19:01 2017
E-mail: Eric2014_Lv@sjtu.edu.cn
@author: DidiLv
"""

import tensorflow as tf
import numpy as np

# define a layer
def add_layer(inputs, in_size, out_size, activation_function = None):
    Weights = tf.Variable(tf.random_normal([in_size,out_size]))
    biases = tf.Variable(tf.zeros([1,out_size]) + 0.1) # better not all zeros
    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

# create data points
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

# placeholder
xs = tf.placeholder(tf.float32, [None, 1])
ys = tf.placeholder(tf.float32, [None, 1])

# structure
l1 = add_layer(xs, 1, 10, activation_function = tf.nn.relu)
prediction = add_layer(l1, 10, 1, activation_function = None)

loss = tf.reduce_mean(tf.reduce_sum(tf.square(ys - prediction), reduction_indices= [1]))

# trian
train_step = tf.train.GradientDescentOptimizer(0.2).minimize(loss)

# initialization
init = tf.global_variables_initializer()

with tf.Session() as sess:
    sess.run(init)
    for i in range(10000):        
        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}))