PATTERN RECOGNITION

Pattern Recognition



Pattern Recognition

Introduction

A.I. An abbreviation for Articial Intelligence, this term has become a very familiar one to many people. To date, many intelligent systems have been developed for a variety of uses, from playing strategic games such as Chess and Warcraft, to assisting with scientism studies, to visual recognition and language processing. The human brain is the most powerful intelligent system currently known to mankind in terms of learning capability, higher reasoning and in the understanding of abstract concepts. It is a system found in every human being and has been in existence for thousands of years, aiding each new generation of humanity tonally result in the modern civilization it is today. Given the power of the human brain, what we propose to do is to attempt to create an AI model based on the human one in the quest for a better AI. We intend to focus on various aspects of the human `A.I.' and also examine some of the other A.I. models currently in use today. We will examine the human learning process, focusing on two important areas: The understanding of abstract concepts and the abilities which are inherent in humans from birth. We will also touch on the possibility of there being more than one processing pathway in the human brain, each used when needed. The theories proposed here will aid in the construction of the human A.I. model.

Literature review

Pattern recognition is the most important aspect of human learning. Pattern recognition is the science that deals with the description and classification (recognition) of objects, people, signs, representations, etc. This science works based on a previously established set of all objects (patterns) to recognize individual. The range of pattern recognition applications is very broad, however the most important are related to vision and hearing by a machine, similar to humans. The outline of a pattern recognition system consists of several interrelated steps (the results of a phase can modify the parameters of previous steps). It shows a general outline of a pattern recognition system, in which the sensor is intended to provide a feasible representation of the elements of the universe to be classified. Sub-system is a crucial because it determines the limits on the performance of the entire system. Feature extraction is the stage that is responsible, from the pattern of representation, removing the discriminatory information eliminating redundant and irrelevant information. The binder is the stage of decision making in the system. Its role is to allocate the patterns of unknown class to the appropriate category. (Nilsson, 1998, p.50)

General Outline of a Pattern Recognition System

The aim of these steps is to adjust the system to be able to classify input signals or objects in one of the predefined classes. To do this you must analyze a number of features and to classify input signals successfully, you need a learning process in which the system creates a model of each of the classes from a training sequence or set of vectors characteristics of each ...