Features Selection and k-NN Parameters Optimization based on Genetic Algorithm for Medical Datasets Classification

Rizki Tri Prasetio, Ali Akbar Rismayadi, Nana Suryana, Rochmanijar Setiady

2018

Abstract

Medical dataset classification is a major data mining problem being researched about for a decade. Most classifiers are designed to learn from the data itself through training process, because expert knowledge to determine classifier parameters is difficult. This research proposes a methodology based on data mining paradigm. This paradigm integrates the search heuristic that is inspired by natural evolution called genetic algorithm with the simplest and the most used learning algorithm, k-nearest Neighbors. The genetic algorithm is used for feature selection and parameter optimization while k-nearest Neighbors is used as a classifier. The proposed method is experimented on five medical datasets of the UCI Machine Learning Repository and compared with original k-NN and other feature selection algorithm i.e., forward selection, backward elimination and greedy feature selection. Experiment results show that the proposed method is able to achieve good performance with significant improvement with p value of t-Test is 0.0011.

Paper Citation

in Harvard Style

Tri Prasetio R., Akbar Rismayadi A., Suryana N. and Setiady R. (2018). Features Selection and k-NN Parameters Optimization based on Genetic Algorithm for Medical Datasets Classification.In Proceedings of the 1st International Conference on Recent Innovations - Volume 1: ICRI, ISBN 978-989-758-458-9, pages 3080-3086. DOI: 10.5220/0009947130803086

in Bibtex Style

@conference{icri18,
author={Rizki Tri Prasetio and Ali Akbar Rismayadi and Nana Suryana and Rochmanijar Setiady},
title={Features Selection and k-NN Parameters Optimization based on Genetic Algorithm for Medical Datasets Classification},
booktitle={Proceedings of the 1st International Conference on Recent Innovations - Volume 1: ICRI,},
year={2018},
pages={3080-3086},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0009947130803086},
isbn={978-989-758-458-9},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Recent Innovations - Volume 1: ICRI,
TI - Features Selection and k-NN Parameters Optimization based on Genetic Algorithm for Medical Datasets Classification
SN - 978-989-758-458-9
AU - Tri Prasetio R.
AU - Akbar Rismayadi A.
AU - Suryana N.
AU - Setiady R.
PY - 2018
SP - 3080
EP - 3086
DO - 10.5220/0009947130803086