机器学习（python+numpy+tensorflow）实战样例

import numpy as np
import tensorflow as tf
import matplotlib.pyplot as plt
ses = tf.Session()
# 构建数据集
np.random.seed(5)
x_data = np.linspace(-1, 1, 100)
y_data = 2 * x_data + 1.0 + np.random.randn(x_data.shape[0]) * 0.4 # np.random.randn 用于增加噪声
# 画图
plt.scatter(x_data, y_data)
plt.plot(x_data, 1.0 + 2 * x_data, color="red", linewidth=3)
plt.show()

# 构建模型
def model(x, w, b):
    return tf.multiply(x, w) + b
x = tf.placeholder("float", name="x")
y = tf.placeholder("float", name="y")
w = tf.Variable(1.0, name="w")
b = tf.Variable(0.0, name="b")
pred = model(x, w, b)

# 训练模型
train_epochs = 10
learning_rate = 0.05
# 均方差
loss_function = tf.reduce_mean(tf.square(y-pred))
# 梯度下降优化器
optimizer = tf.train.GradientDescentOptimizer(learning_rate).minimize(loss_function)
ses.run(tf.global_variables_initializer())
for _ in range(10):
    for xs, ys in zip(x_data, y_data):
        loss, _ = ses.run([loss_function, optimizer], feed_dict={x: xs, y: ys})
    btmp = b.eval(ses)
    wtmp = w.eval(ses)
    print("w=", wtmp, ", b=", btmp, "loss=", loss)

ses.close()

步骤：

1. 生成人工数据集及可视化（matplotlib）

2. 构建线性模型

3. 定义损失函数

4. 定义优化器，最小化损失函数

5. 训练结果的可视化

6. 利用学习到的模型进行预测