An Efficient Technique for Protein Sequence Classification Using Data Mining

Abstract

Various classification techniques have been developed for the classification of protein sequences using feature selection method. Feature selection is important in accurate classification. This paper discussed some popular methods that involve feature selection for precise classification of protein sequences such as neural network based classifier, fuzzy method based classifier and rough set based classifier with their respective accuracy and drawback. Feature selection method is used for accurate classification of protein sequence. A new classification technique is proposed here following these discussed methods. The newness of the model proposed here is the aggregation of using intelligent method and introduction of a new technique for selecting specific features to classify protein sequence accurately and faster. Fuzzy classifier, neural method and rough set classifier are aggregated in a single model. The method’s primary aim is to identify and classify the protein sequences based on extracting definite features from each sequence very fast with maintaining the accuracy level. Use of a new technique for extracting specific features reduces the computational overhead effectively. Comparing with the pervious methods a Great reduction of execution time without affecting accuracy level is achieved.