Where to start Handwritten Text Recognition using Neural Network in C# or any language

I am doing project on hand written text recognition systems. I need help where to开发者_如何学C start.......

If I were interested in a project like that I think I would like...I dunno...go to the library and get some books or something.

There's a good bit on information on this topic here http://www.codeproject.com/KB/library/NeuralNetRecognition.aspx and there's also a paper on the topic here http://www.cs.utoronto.ca/~hinton/absps/ijcai05.pdf . This isn't really my area, but honestly your question is really open-ended and kind of hard to answer. This is a fairly large topic, however, so you should be able to find something in university papers like the one I linked or in any discussions of currently implemented systems like Palm's, for example.

Start with understanding of "What is a neural network" ?

What are its various types and properties of each network type

Then, learn about feedback mechanism and understand error propagation and learning mechanism through hidden layers

Then create a tree - based data structure in C/C++ and try to have network representation in your model

http://www.learnartificialneuralnetworks.com/