tensorflow之非线性方程

[python]
import tensorflow as tf
import numpy as np
import matplotlib.pyplot as plt


def add_layer(inputs, in_size, out_size, activation_function=None):  # activation_function=None线性函数
    Weights = tf.Variable(tf.random_normal([in_size, out_size]))  # Weight中都是随机变量
    biases = tf.Variable(tf.zeros([1, out_size]) + 0.1)  # biases推荐初始值不为0
    Wx_plus_b = tf.matmul(inputs, Weights) + biases  # inputs*Weight+biases
    if activation_function is None:  # 线性回归
        outputs = Wx_plus_b
    else:  # 逻辑回归
        outputs = activation_function(Wx_plus_b)
    return outputs


def get_next_batch(batch_size=64):
    x = np.random.choice(np.linspace(start=0, stop=1, num=200), batch_size)  # 从0-1之间选取batch_size个数,免去对数据的归一化
    x = np.reshape(x, newshape=[-1, 1])
    # noise=np.random.normal(0, 0.05, x.shape)*50 # 噪声
    y = np.power(x, 2)
    return x, y


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


def model():    
    layer_1 = add_layer(x_input, 1, 4, activation_function=tf.nn.relu)  # 隐藏层
    layer_2 = add_layer(layer_1, 4, 4, activation_function=tf.nn.relu)  # 隐藏层
    prediction = add_layer(layer_2, 4, 1, activation_function=None)  # 输出层
    return prediction


model_dir = 'model_non_linear'  # 模型存储目录


def train():
    prediction = model()
    # 求标准差
    loss = tf.reduce_mean(
        tf.reduce_sum(tf.square(y_input - prediction), reduction_indices=[1]))  # square()平方,sum()求和,mean()平均值

    optimizer = tf.train.GradientDescentOptimizer(0.005).minimize(loss)  # 0.05学习效率,minimize(loss)减小loss误差

    sess = tf.Session()
    saver = tf.train.Saver()
    checkpoint = tf.train.latest_checkpoint(model_dir)
    if checkpoint:
        saver.restore(sess, tf.train.latest_checkpoint(model_dir))  # 从模型中读取参数
    else:
        sess.run(tf.global_variables_initializer())  # 变量初始化

    # 训练20万次
    for step in range(200000):
        x_train, y_train = get_next_batch(batch_size=256)
        loss_, optimizer_ = sess.run([loss, optimizer], feed_dict={x_input: x_train, y_input: y_train})
        if step % 50 == 0:
            print(step, loss_)
    # 模型保存
    saver.save(sess, '{}/non_linear_equation'.format(model_dir), global_step=step)


def test():
    predition = model()
    sess = tf.Session()
    saver = tf.train.Saver()
    saver.restore(sess, tf.train.latest_checkpoint(model_dir))
    x_test, y_test = get_next_batch(200)

    y_pred = sess.run(predition, feed_dict={x_input: x_test})

    # 数据可视化
    fig = plt.figure()
    ax = fig.add_subplot(2, 1, 1)
    ax.scatter(x_test, y_test)

    bx = fig.add_subplot(2, 1, 2)
    bx.scatter(x_test, y_pred, c='r')
    plt.xlabel('X')
    plt.ylabel('Y')

    plt.show()


if __name__ == '__main__':
    train()
    # test()

训练20万次后得到下图