PLA(Regression)

Type Of this Sample

Output Space Data Label Protocol Input Space
Regression Supervised Online Raw

該範例沒有使用Activation Function 所以只能學習線性

Backpropagation

Implement PLA(Regression)

資料=10x+5 嘗試要讓類神經藉由數據x跟y學到回歸直線 學習效率：0.0000001(讓速度慢一點好顯示)

Code
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
import  tensorflow as tf
import numpy as np
import matplotlib.pyplot as plt

x_data=np.linspace(1,11,111).astype(np.float32)
#x_data=np.random.rand(100).astype(np.float32)
y_data=10*x_data+5


Weights=tf.Variable(tf.zeros([1]))

biases=tf.Variable(tf.zeros([1]))


y=Weights*x_data+biases

#loss = tf.reduce_mean(tf.square(y-y_data))#用最小平方法 可以求出回歸直線 不能亂用

loss=tf.reduce_mean(tf.reduce_sum(tf.square(y-y_data)))
#************Optimizer*********
#****MomentumOptimizer
#optimizer=tf.train.MomentumOptimizer(0.1,0.2)#梯度下降法 Gradient descent


optimizer=tf.train.GradientDescentOptimizer(0.0000001)


train =optimizer.minimize(loss)

init=tf.global_variables_initializer()

sess=tf.Session()
sess.run(init)

#*********** for plt
fig = plt.figure()
ax = fig.add_subplot(1,1,1)
ax.scatter(x_data, y_data,color="green")
lines = ax.plot(3, 3, 'r-', lw=1)
ax.legend(labels=['prediction line','y=10x+5'],loc='best')
plt.ion()
plt.show()
plt.pause(2.5)
for step in range(20000):
    if step % 100 == 0:
        # 只是顯示參數
        try:
            ax.lines.remove(lines[0])
        except Exception:
            pass
        print(step, sess.run(Weights), sess.run(biases), sess.run(loss))
        #for plt

        lines = ax.plot(x_data, sess.run(y), 'r-', lw=1)
        plt.pause(1)


    sess.run(train)  # 真正訓練

評論