Skip to Main Content
The high-dimensional feature vectors often impose a high computational cost when classification is performed. Feature selection plays major role as a pre-processing technique in reducing the dimensionality of the datasets in data analysis and data mining. This process reduces the number of features by removing irrelevant and redundant data and hence resulting in acceptable classification accuracy. Filter and wrapper are the two kinds of feature selection methods. Experimental results have proved that the wrapper methods can yield better performance, although they have the disadvantage of high computational cost. This paper presents a Harmony Search based novel optimization algorithm for wrapper feature selection. 1-NN classifier method has been used to evaluate the quality of the solutions. The performance of the proposed approach has been analysed by experiments with various real-world data sets. The proposed method, HS-1-NN, produced better performance than other state-of-the-art methods in terms of classification accuracy and convergence rate.