-
Notifications
You must be signed in to change notification settings - Fork 1
/
svm.py
45 lines (33 loc) · 1.04 KB
/
svm.py
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
import numpy as np
x1 = np.array([2.327868056, 3.032830419, 4.485465382, 3.684815246, 2.283558563, 7.807521179, 6.132998106, 7.514829366, 5.502385039, 7.43293235])
x2 = np.array([2.458016525, 3.170770366, 3.696728111, 3.846846973, 1.853215097, 3.290132136, 2.140563087, 2.107056961, 1.404002608, 4.236232628])
y = np.array([-1, -1, -1, -1, -1, 1, 1, 1, 1, 1])
L=0.45
b1=0.0
b2=0.0
t=1
epochs=16
for i in range(epochs):
for i in range(10):
output=y[i]*(b1*x1[i])+(b2*x2[i])
if output<1:
b1=((1-1/t)*b1)+(1/(L*t))*(y[i]*x1[i])
b2=((1-1/t)*b2)+(1/(L*t))*(y[i]*x2[i])
else:
b1=(1-1/t)*b1
b2=(1-1/t)*b2
t=t+1
print("b1",b1)
print("b2",b2)
print(b1,b2)
for i in range(len(x1)):
yp = b1*x1[i] + b2*x2[i]
print("Y predicted :", yp)
if(yp<0):
print("-1")
y_p= -1
else:
print("1")
y_p = 1
error=y[i]-y_p
print("Total Error is :", error)