Simulated Relief Algorithm for Feature Selection Approach

Feature selection is an important preprocessing technique in data mining and it is the process of selecting the relevant features from the data sets. The objective of the feature selection techniques are to reduce the number of features and to improve the classification accuracy. Three contributions such as sequential backward selection algorithm, Relief algorithm and simulated annealing algorithm are combined and a new novel algorithm known as Simulated Relief is proposed in this paper. The efficiency and the effectiveness of the proposed algorithm is evaluated with cotton data set provided by central cotton Research station at Coimbatore. Weka tool and Microsoft excel sheet contribute the data manipulation task for computation process. The experimental study concludes that the Simulated Relief algorithm using Multilayer perceptron classifier provides higher classification accuracy than using the Nave bayes classifier.