High Order Neural Networks for Pattern Recognition

This thesis is concerned with study of high order neural networks for character recognition. The goal of the thesis is to verify and test models of high order neural networks, which are applicable to character recognition with regard to various deformations (rotation, scaling, translation ...). Our modi cations of these models are designed to improve classi cation abilities of the models. Obtained experimental results are summarized and compared with each other. The thesis also includes an evaluation of the results