Avik Samanta¹ and Suprativ Saha², ¹JIS College of Engineering, India and ²Global Institute of Management and Technology, India

Feature selection is more accurate technique in protein sequence classification. Researchers apply some well-known classification techniques like neural networks, Genetic algorithm, Fuzzy ARTMAP, Rough Set Classifier etc for extracting features.This paper presents a review is with three different classification models such as fuzzy ARTMAP model, neural network model and Rough set classifier model.This is followed by a new technique for classifying protein sequences.The proposed model is typically implemented with an own designed tool using JAVA and tries to prove that it reduce the computational overheads encountered by earlier approaches and also increase the accuracy of classification.

Data Mining, Neural Network Model, Fuzzy ARTMAP Model, Rough Set Classifier, Protein Sequence, 2-gram encoding method, 6-letter exchange group method.