Citation

Wong, Ee Thiing and Ong, Lee Yeng and Leow, Meng Chew and Tan, Joon Liang and Wong, Johnny Stanley and Gan, Xiu Lin (2024) Experimental Study on Feature Selection Methods Applied to Type-2 Diabetes Microbiome Data. In: 2024 12th International Conference on Information and Communication Technology (ICoICT), 07-08 August 2024, Bandung, Indonesia.

Text Experimental Study on Feature Selection Methods Applied to Type-2 Diabetes Microbiome Data.pdf - Published Version Restricted to Repository staff only Download (736kB)

Abstract

—Due to the reduction in the cost of shotgun sequencing and the improvement in resolution, metagenomic datasets are anticipated to reveal a greater relevance on the roles of microbiome in the human body. Hence, it is believed that microbial species can be utilized as a marker in clinical settings to understand, diagnose, and treat Type-2 Diabetes (T2D). However, despite the importance of feature selection methods in analyzing microbiome data, there is a lack of comparative studies specifically focused on T2D. To address this gap, our study aims to explore and compare commonly used feature selection models applied to T2D microbiome data. This study aims to identify potential gut microbiome species with high correlation to T2D via these feature selection methods. This paper focuses on the discussion of pre-processing steps and the evaluation of feature selection methods which potentially lead to the discovery of T2D marker species. The feature selection methods that are being compared in this study include Recursive Feature Elimination (RFE), Chi-Squared, Random Forest Feature Importance and Information Gain. The identified species are Bacteroides uniformis. Faecalibacterium prausnitzii, Flavonifractor plautii, Lachnospiraceaen bacterium and Escherichia coli. Other than identifying the potential marker species of T2D, this study highlights the strengths and limitations of various feature selection methods, providing insights for future research and clinical applications

Item Type:	Conference or Workshop Item (Paper)
Uncontrolled Keywords:	Diabetes , T2D , recursive feature selection , chi-squared , information gain , random forest feature importance
Subjects:	Q Science > QA Mathematics > QA71-90 Instruments and machines R Medicine > RA Public aspects of medicine > RA421-790.95 Public health. Hygiene. Preventive medicine
Divisions:	Faculty of Information Science and Technology (FIST)
Depositing User:	Ms Nurul Iqtiani Ahmad
Date Deposited:	03 Dec 2024 03:38
Last Modified:	03 Dec 2024 03:39
URII:	http://shdl.mmu.edu.my/id/eprint/13187

Downloads

Edit (login required)