Comparative Study & Performance Evaluation of Various Classifiers Using a Data Set

抄録

Real-world knowledge discovery processes typically consist of complex data pre-processing, machine learning, evaluation, and visualization steps. Hence a data mining platform should allow complex nested operator chains or trees, provide transparent data handling, comfortable parameter handling and optimization, be flexible, extendable and easy-to-use. Modern machine learning techniques have encouraged interest in the development of various systems that ensure secure, reliable and many more operations in the different fields and applications. In an earlier study, many other approaches/methods were investigated to develop various applications using modern machine learning techniques and more specific classification algorithms. The Weka machine learning workbench provides a general-purpose environment for automatic classification, clustering and feature selection, and common data mining problems in bioinformatics research. Here in this Project Report paper we have used various classifiers with filters to perform classification and we have done analysis of data with different classifiers and then we have done feature selection process and during all these activities we have observed and record the various performance change and different graphs which are briefed inside this paper.