Machine Learning and Big Data

6Pattern Recognition Concepts

Ambeshwar Kumar, R. Manikandan* and C. Thaventhiran

School of Computing SASTRA Deemed University, Thanjavur, India

Abstract

The area of pattern recognition is concerned with the discovery of regularities in data by using machine learning algorithms. Make use of these regularities patterns to take an action for classifying the object in different categories. A pattern might be physically visible or it can be derived mathematically using algorithms. Consider an example to determine the two categories of fish like rohu and catla both looks similar but its pattern gives an idea to classify (i.e., head pattern of those fishes were different). The small changes in pattern are not able to classify easily and then it solved by using algorithms. In biomedical, image classification task to distinguish the matching pattern image and unmatched pattern is handled by pattern recognition techniques. Pattern recognition gives better ideas to further development of computational technology, with the help of machine learning and big data analytics to make a smarter system. As per the security purpose, it is more essential to track the forgery in bank, biometric tracking, and left luggage at the airport; these can be done easily by pattern recognition techniques. Learning a new knowledge or experience and linked it with the old or existing knowledge, sometimes, it might be new patterns got generated. Consciously learning patterns merged with the newly generated patterns ...