Feature Selection in Machine Learning: Methods and Comparison

Abstract

Nowadays, a huge amount of data is generated every day in continuous manner in every hour and if the data is not utilized in the right or meaningful manner then this is just like garbage. Therefore, the meaningful information from the data can be represented through the feature selection. Feature selection refers to various methods which selects the most appropriate from the data according to the problem. This paper provides an insight into various feature selection methods which include filter, wrapper, embedded, hybrid and a detailed explanation of various techniques utilized by these methods. Further, a comparison among these feature selection methods have also been made.