A Survey on the Use of Pattern Recognition Techniques

Abstract

Pattern recognition is an innate cognitive process of matching information from the environment with the information stored in memory. Core methods are successful in many areas of numerical analysis, pattern recognition and machine learning. These are methods which generate an abstracting model from given observations (objects, measurements) in a training step, which subsequently allows generalizing statements for new observations. Various approaches are used to implement a pattern recognition system. In this paper we will discuss Statistical, Structural, hybrid and Neural Network based approach.