Enhancing classification accuracy by attribute reduction technique

Abstract

Data mining methods helps in analyzing the data set efficiently by reducing the size of the search space, so as to choose significant attribute for recognition of the type of appropriate data. This study deals with the classification of categories of glass, which helps in criminology investigation. The glass material got as evidence in the crime scene is to be correctly identified. The fundamental purpose of this work is to deal with a large glass data set with high accuracy of identifying king of the glass. The models are constructed with supervised learning algorithm in weka tool. It is important to minimize the dimensions of data by constructing the models by selecting the attributes that is implemented as search methods, which are applied to predict the evaluating for the possible test cases.