tensorflow第二面非线性回归预测

In [17]:

# note from dataguru Tensorflow神经网络框架-12周(2017全新)
import tensorflow as tf
import numpy as np
import matplotlib.pyplot as plt
x_data = np.linspace(-0.5,0.5,200)[:,np.newaxis]
noise = np.random.normal(0,0.02,x_data.shape)
y_data = np.square(x_data) + noise

x = tf.placeholder(np.float32,[None,1])
y = tf.placeholder(np.float32,[None,1])
                   
#middle layer 1
Weights_L1 = tf.Variable(tf.random_normal([1,10]))
bias_L1 = tf.Variable(tf.zeros([1,10]))
Wx_bias_L1 = tf.matmul(x,Weights_L1) + bias_L1
L1 = tf.nn.tanh(Wx_bias_L1)

#output layer
Weights_L2 = tf.Variable(tf.random_normal([10,1]))
bias_L2 = tf.Variable(tf.zeros([1,1]))
Wx_bias_L2 = tf.matmul(L1,Weights_L2) + bias_L2
prediction = tf.nn.tanh(Wx_bias_L2)

#损失函数
loss = tf.reduce_mean(tf.square(y-prediction))
#梯度下降, 下降步长0.01
train_step = tf.train.GradientDescentOptimizer(0.1).minimize(loss)

with tf.Session() as sess:
    sess.run(tf.global_variables_initializer())
    for _ in range(1,2000):
        sess.run(train_step,feed_dict={x:x_data, y:y_data})
        
    #预测一下
    pred_value = sess.run(prediction,feed_dict={x:x_data})
    
    #showing
    plt.figure()
    plt.scatter(x_data,y_data)
    plt.plot(x_data,pred_value,'r-',lw=5)
    plt.show()

Output: