Neural-Network-Algorithms/madaline.m at master · 3point141/Neural-Network-Algorithms · GitHub

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
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
%author : aman
%for the XOR implementation, bipolar input and bipolar output
% x1=[1 1 -1 -1]
% x2=[1 -1 1 -1]
% t=[-1 1 1 -1]
% w1=[0.05 0.2]
% b1=0.3
% w2[0.1 0.2]
% b2=0.15
% w3=[0.5 0.5]
% b3=0.5
% alpha=0.5
% epoch=1

clc
close all
clear all

x1=input('enter the first input matrix : ');
x2=input('enter the second input martix : ');
t=input('enter the target of these matrix : ');

w1=input('enter the weights into z1 ( w11,w21,w31,w41...) : ');
b1=input('enter the first bias (b1) : ');
w2=input('enter the weights into z2 ( w12,w22,w32,w42...) : ');
b2=input('enter the second bias (b2) : ');
w3=input('enter the weights into z3 ( v1,v2,v3,v4...) : ');
b3=input('enter the third bias (b3) : ');
alpha=input('enter the learning rate : ')
epoch=input('enter the number of epoch : ')
l=length(x1);
pp=1;
while pp<=epoch
    for i=1:l

        sum1=0;
        sum2=0;
        k=1;
        sum1=x1(i)*w1(k) + x2(i)*w1(k+1);
        sum2=x1(i)*w2(k) + x2(i)*w2(k+1);
        x=[x1(i) x2(i)]
        zin1=b1+sum1;
        zin2=b2+sum2;
        xx=zin1;
        yy=zin2;

        if zin1>=0
            zin1=1
        else
            zin1=-1
        end

        if zin2>=0
            zin2=1
        else
            zin2=-1
        end

        yin=(x1(i)*w3(k)+x2(i)*w3(k+1))+b3

        if yin>=0
            yin=1
        else
            yin=-1
        end

        res=t(i)-yin
        if res==0
            disp('no weights will be updated ')
            continue;
        else
            disp('going to update weights ... ')
            if t(i)==1
                if abs(xx)>abs(yy)
                    closest=yy;
                    b2=b2+alpha*(1-yy);
                    disp('t==1 , b2 updated... ')
                    b2
                    le=length(w2);
                    for j=1:le
                        w2(j)=w2(j)+alpha*(1-yy)*x(j);
                    end
                    disp('t==1, w2 updated... ')
                    w2
                else
                    closest=xx;
                    b1=b1+alpha*(1-xx);
                    disp('t==1, b1 updated... ')
                    b1
                    le=length(w1);
                    for j=1:le
                        w1(j)=w1(j)+alpha*(1-xx)*x(j);
                    end
                    disp('t==1, w1 updated : ')
                    w1
                end

            else
                if t(i)==-1
                    if xx>0
                        b1=b1+alpha*(-1-xx);
                        disp('t==-1, b1 updated... ')
                        b1
                        le2=length(w1);
                        for j=1:le2
                            w1(j)=w1(j)+alpha*(-1-xx)*x(j);
                        end
                        disp('t==-1, w1 updated : ')
                        w1
                    end

                    if yy>0
                        b2=b2+alpha*(-1-yy);
                        disp('t==-1, b2 updated.. ')
                        b2
                        le2=length(w2);
                        for j=1:le2
                            w2(j)=w2(j)+alpha*(-1-yy)*x(j);
                        end
                        disp('w2 updated :')
                        w2
                    end

                end

            end

        end
    end
pp=pp+1;
disp('new epoch started ........................................... ')
end

w1
w2
b1
b2