1 % Sensor classification problems. X is the input sample, and T is the target vector.
2 Clear
3 % X = [ - 0.5 , - 0.5 , 0.3 , 0.1 , - 0.1 , 0.8 , 0.2 , 0.3 ;
4 % 0.3 , - 0.2 , - 0.6 , 0.1 , - 0.5 , 1.0 , 0.3 , 0.9 ];
5 % T = [ 0 , 0 , 0 , 1 , 0 , 1 , 1 , 1 ]; M = 100 ;
6 % E = [ 0.01 ; 0.01 ; 0.0 ; 0.0 ; 0.01 ; 0.01 ; 0.01 ; 0.01 ];
7
8
9 % Logical or classification instance
10 X = [ 0 , 0 , 1 , 1 ;
11 0 , 1 , 0 , 1 ];
12 T = [ 0 , 1 , 1 , 1 ];
13 M = 100 ;
14 E = [ 0.1 0.1 0.1 0.1 ] ';
15
16 % Classification of two types of specimens by single neuron structure Sensor
17 [R, Q] = Size (X ); % R = 2 ; Q = 8 is the dimension of array X;
18 [S, q] = Size (t ); % S = 1 ; Q = 8 is the dimension of the T array;
19 W = Rands (S, R ); % Generates initial values of random weight array W based on the dimensions of input sample X and target vector T;
20 % W = [ - 0.1 0.2 ];
21 C = Rands (S, 1 ); % Generate initial Random threshold values;
22 Y = W * X + C; % Computing sensor output;
23 F = Zeros ([S q]); % Initialize and allocate memory
24 For J = 1 : M % Cycle start
25 For I = 1 : Q % Classification of sensor output;
26 If Y ( 1 , I) > 0 ; % Evaluate the output after the sensor threshold value is triggered;
27 F ( 1 , I) = 1 ;
28 Else
29 F ( 1 , I) = 0 ;
30 End
31 End
32 If All (F = T) % The weight array and threshold are corrected when the output values of sensor F and target t are not equal;
33 Break
34 End
35 DC = 0.1 * (T -F ) * E; % Threshold Value correction;
36 C = C + DC; % Threshold Value correction;
37 DW = (T -F ) * X '; % Weight increment
38 W = W + DW; % Dynamic Weight
39 Y = W * X + C;
40 End % Loop ends
41 Plotpv (x, t ); % Painting point, drawing classification Mode
42 Plotpc (W, C ); % Draw the dividing line and draw the decision Surface
43 Grid on
44 Xlabel ('x1 '), ylabel ('x2 ');
45 Title (sprintf ('( % F) x1 + ( % F) x2 + ( % F) ', w ( 1 ), W ( 2 ), C ));
46
