A Correlation - Sequential Forward Selection Based Feature Selection Method for Healthcare Data Analysis

Abstract

In recent years healthcare data is growing exponentially and thus contains a huge number of attributes. The major challenge is to predict and analysis of all these data in a huge format. Feature selection is a solution in which a subset of informative features is selected from a huge high dimensional dataset and used for analysis. Feature Selection helps to increase accuracy and remove all irrelevant features. Filter, Wrapper, and Embedded are different methods for feature selection. This paper presents a correlation-sequential forward selection based hybrid feature selection algorithm on a healthcare dataset and analyses result obtained after applying different machine learning algorithm on the feature subset. Correlation-based feature selection is part of the filter method and sequential forward selection is based on the wrapper method. Selecting important attributes for healthcare is essential as it has a direct effect on human health. Furthermore, this paper is an attempt to make healthcare predictions more accurately and in a timely manner. Correlation-based feature selection is part of the filter method and sequential forward selection is based on the wrapper method. Selecting important attributes for healthcare is essential as it has a direct effect on human health. Furthermore, this paper is an attempt to make healthcare predictions more accurately and in a timely manner.