Comparative study of feature selection methods to analyze performance of lung cancer data

No Thumbnail Available

Date

2015

Journal Title

Journal ISSN

Volume Title

Publisher

IADIS

Open Access Color

OpenAIRE Downloads

OpenAIRE Views

Research Projects

Organizational Units

Journal Issue

Abstract

Feature selection, also known as attribute selection, is a process which attempts to select more informative features among datasets to be used in model construction. The main aim of feature selection can improve the prediction accuracy and reduce the computational overhead of classification algorithms. In this study, several approaches such as Information Gain Attribute Evaluation, Chi-Squared Attribute Evaluation, Filtered Attribute Evaluation, Gain Ratio Attribute Evaluation and Symmetrical Uncertainty Attribute Evaluation are carried out to discover the discriminative features on the same disease, namely lung cancer, using four different medical datasets. The efficiency of each approach is evaluated using machine learning software.

Description

Keywords

Chi-squared attribute evaluation, Feature selection, Filtered attribute evaluation, Gain ratio attribute evaluation, Information gain attribute evaluation, Symmetrical uncertainty attribute evaluation

Turkish CoHE Thesis Center URL

Fields of Science

Citation

0

WoS Q

Scopus Q

Source

Proceedings of the European Conference on Data Mining 2015, ECDM 2015 and International Conferences on Intelligent Systems and Agents 2015, ISA 2015 and Theory and Practice in Modern Computing 2015, TPMC 2015 - Part of the Multi Conference on Computer Science and Information Systems 2015 -- European Conference on Data Mining 2015, ECDM 2015 and International Conferences on Intelligent Systems and Agents 2015, ISA 2015 and Theory and Practice in Modern Computing 2015, TPMC 2015 -- 22 July 2015 through 24 July 2015 -- Las Palmas de Gran Canaria -- 119133

Volume

Issue

Start Page

219

End Page

222