Pattern Classification Using Optimized Machine Learning Techniques

Abstract

Most of the real world problems in engineering, medicine, industry, science and business also involve data classification. Classification is a supervised machine learning technique used to predict group membership for data instances. Pattern classification problems belong to the category of supervised learning. Pattern Classification involves assigning a label to a given input data. Neural Networks are an effective tool in the field of pattern classification, using training and testing data to build a model. Training neural networks in classification problems, especially when biological data are is a very challenging task. The protein superfamily classification problem, which consists of determining the superfamily membership of a given unknown protein sequence, is very important for a biologist for many practical reasons, such as drug discovery, prediction of molecular function and medical diagnosis. The objective of this work is creating a classification model for classifying data using Multilayer feed forward network. It contain two phases. First, classifier model was build for iris plant classification. Second, classifier model was build for protein sequence classification to know the organism of protein and family of the given protein sequence.