TensorFlow学习笔记（十）tf搭建神经网络可视化结果

 代码

"""

Please note, this code is only for python 3+. If you are using python 2+, please modify the code accordingly.

"""

import tensorflow as tf

import numpy as np

import matplotlib.pyplot as plt

def add_layer(inputs, in_size, out_size, activation_function=None):

    # add one more layer and return the output of this layer

    Weights = tf.Variable(tf.random_normal([in_size, out_size]))

    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

# Make up some real data

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

# define placeholder for inputs to network

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

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

# add hidden layer

l1 = add_layer(xs, 1, 10, activation_function=tf.nn.relu)

# add output layer

prediction = add_layer(l1, 10, 1, activation_function=None)

# the error between prediciton and real data

loss = tf.reduce_mean(tf.reduce_sum(tf.square(ys - prediction),

                     reduction_indices=[1]))

train_step = tf.train.GradientDescentOptimizer(0.1).minimize(loss)

# important step

#init = tf.initialize_all_variables()

init = tf.global_variables_initializer()

sess = tf.Session()

sess.run(init)

# plot the real data

fig = plt.figure()

ax = fig.add_subplot(1,1,1)

ax.scatter(x_data, y_data)

plt.ion()

plt.show()

for i in range(1000):

    # training

    sess.run(train_step, feed_dict={xs: x_data, ys: y_data})

    if i % 50 == 0:

        # to visualize the result and improvement

        try:

            ax.lines.remove(lines[0])

        except Exception:

            pass

        prediction_value = sess.run(prediction, feed_dict={xs: x_data})

        # plot the prediction

        lines = ax.plot(x_data, prediction_value, 'r-', lw=5)

        plt.pause(0.1)

结果：