import tensorflow as tf import numpy as np import matplotlib.pyplot as plt if __name__=='__main__': x_data=np.linspace(-1,1,200)[:,np.newaxis] #构建等差数列，:表示构建的等差数列，newaxis，创建新列，及把一维变2维 noise=np.random.normal(0,0.2,x_data.shape) y_data=np.square(x_data)+noise f1=plt.figure(1) plt.subplot(211) plt.scatter(x_data,y_data,marker='o',color='r',label='1',s=5) plt.show() x=tf.placeholder(tf.float32,[None,1]) y = tf.placeholder (tf.float32, [None, 1]) Weight_L1=tf.Variable(tf.random_normal([1,10])) biases_L1=tf.Variable(tf.zeros([1,10])) Wx_plus_L1=tf.matmul(x,Weight_L1)+biases_L1 L1=tf.nn.tanh(Wx_plus_L1) Weight_L2=tf.Variable(tf.random_normal([10,1])) biases_L2=tf.Variable(tf.zeros([1,1])) Wx_plus_L2=tf.matmul(L1,Weight_L2)+biases_L2 predication=tf.nn.tanh(Wx_plus_L2) #二次代价函数 loss=tf.reduce_mean(tf.square(y-predication)) #梯度下降法 train_step=tf.train.GradientDescentOptimizer(0.1).minimize(loss) with tf.Session()as sess: sess.run(tf.global_variables_initializer()) for _ in range(2000): sess.run(train_step,feed_dict={x:x_data,y:y_data}) #获取测试值 pre_val=sess.run(predication,feed_dict={x:x_data}) #画图 plt.figure(1) plt.scatter(x_data,y_data) plt.plot(x_data,pre_val,'b-',lw=2) plt.show()
基础功能，搭建一个神经网络并画图，作为对tensorflow的了解格式转换：jupyter nbconvert --to script test.ipynb 将ipynb转py
;