A Meta-Heuristic Approach to Feature Selection using Harmony Search

Main Article Content

D. Sai Kumar
D. Dedeepya


Most of the real world applications deal with large amounts of data that may be in Gigabytes or Terabytes. When we come to analyze this type of data it is not so easy, some practical problems arise (i.e. curse of dimensionality) , so here we use a concept called Feature Selection. The main aim of the Feature Selection is to discover a minimal feature subset from a problem domain, while retaining a suitably high accuracy in representing the original data. Many search strategies have been exploited for the task of Feature Selection, in an effort to identify more compact and better quality subsets. In this work, a novel FS approach based on harmony search (HS) is presented. Harmony Search is a recently developed meta-heuristic algorithm that mimics the improvisation process of a music player. Each musician plays a note while finding best notes of Harmony altogether. The simplicity of the Harmony Search is exploited to reduce overall complexity of search process. This work has described a flexible Feature Selection method based on general Harmony Search (HS). The thesis shows that the Harmony Search is capable of identifying better-quality feature subsets for most data sets than correlation feature selection (cfs) subset evaluator and consistency based subset evaluator.

Keywords: Feature Selection, Harmony Search, Cfs, Subset evaluator


Download data is not yet available.

Article Details