Skip to content

Latest commit

 

History

History
30 lines (22 loc) · 891 Bytes

File metadata and controls

30 lines (22 loc) · 891 Bytes

Implementation-of-Hopfield-neural-network-for-Character-Recognition

Goal: Recognition of six first uppercase English letters by Hopfield neural network (even if they are noisy)

training set: ABCDEFZ

test set: noisy images

STEPS:

1. Create Network:

[handles.X handles.T]=prprob();
handles.Letters='ABCDEFZ';
handles.LettersIndex=double(handles.Letters)-64;
handles.P=handles.X(:,handles.LettersIndex);
handles.net=newhop(handles.P);

image

2. Add Noise:

a = a + 0.1 * randn(size(a));

image

3. Test and RUN!

image