i have an assignment on pattern recognition,to train a Hebb neural network to recognize different 26 (upper case) letters of the alphabet,only one output representing the correct letter should be ‘active’ at a time in matlab. i have entered all my patterns (A...Z) and the target as well and training it gives the correct output, but when i change one of the pixel (from 1 to 0) of a single letter it still gives me the same out i.e. its not training. What is it that am doing wrong?? am using the neural network GUI in matlab
crazyIT
0
Newbie Poster
Taywin
312
Posting Virtuoso
My first guess that I would look to is how the agent learn a pattern. This also involves how you read data of each letter. When you said "change one of the pixel," it means nothing to anyone because no one knows how a letter data represent but you. Anyway, you may need to create a simple test data to see that your agent actually learns correctly. That should answer most of your questions now...
Be a part of the DaniWeb community
We're a friendly, industry-focused community of developers, IT pros, digital marketers, and technology enthusiasts meeting, networking, learning, and sharing knowledge.