Character recognition using Teaching Learning Based Optimization

Abstract

Character recognition is an image analysis method, where handwritten images are given as input to a system and then the job of the system is to recognize them on the basis of information available about them. Pattern recognition capability of human beings cannot be imitated, however up to a certain extent it can be achieved by the use of neural network. In this paper an attempt is made to recognize English characters by the use of back propagation algorithm (BPA) as well as Teaching Learning Based Optimization (TLBO). TLBO is a recent algorithm used to solve many real world problems, which is inspired by practical environment of a class room, whereas Back Propagation algorithm is a generalization of Least Mean Square algorithm.