Abstract: The huge amount of data is collected from the real world and stored in the databases. The extraction of useful information from the large data is a tedious task. The effective extraction of the data can obtain useful information. The classification of the large data regarding their features is difficult. The feature selection for the particular data set in order to classify the data should be performed in clear way to obtain better classification. Different algorithms are used in the data mining to obtain the classification in a better manner with effective feature selection process. The problem confronted by the classification process is due to the presence of redundant data and irrelevant data. In feature selection process the data present in the set may not support with similar feature as some points having irrelevant feature then specified. The existing classification process can’t compete with the minimum local deviation and they include high computation cost. In order to achieve better classification of the data to obtain the appropriate result this work includes implementation of the Pareto Optimization along with optimization algorithm to extract the feature subset. The Pareto front is used to determine non-dominated feature subset in order to compute the feature extraction in minimal number and to increase the classification accuracy.

Keywords: Classification, Feature Selection, Pareto Optimization, Optimization Algorithm.