A Pareto Front Approach for Feature Selection

Enguerran Grandchamp, Mohamed Abadi, Olivier Alata

2016

Abstract

This article deals with the multi-objective aspect of an hybrid algorithm that we propose to solve the feature subset selection problem. The hybrid aspect is due to the sequence of a filter and a wrapper method. The filter method reduces the exploration space by keeping subsets having good internal properties and the wrapper method chooses among the remaining subsets with a classification performances criterion. In the filter step, the subsets are evaluated in a multi-objective way to ensure diversity within the subsets. The evaluation is based on the mutual information to estimate the dependency between features and classes and the redundancy between features within the same subset. We kept the non-dominated (Pareto optimal) subsets for the second step. In the wrapper step, the selection is made according to the stability of the subsets regarding classification performances during learning stage on a set of classifiers to avoid the specialization of the selected subsets for a given classifiers. The proposed hybrid approach is experimented on a variety of reference data sets and compared to the classical feature selection methods FSDD and mRMR. The resulting algorithm outperforms these algorithms.

Paper Citation

in Harvard Style

Grandchamp E., Abadi M. and Alata O. (2016). A Pareto Front Approach for Feature Selection . In Proceedings of the 5th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM, ISBN 978-989-758-173-1, pages 334-342. DOI: 10.5220/0005752603340342

in Bibtex Style

@conference{icpram16,
author={Enguerran Grandchamp and Mohamed Abadi and Olivier Alata},
title={A Pareto Front Approach for Feature Selection},
booktitle={Proceedings of the 5th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM,},
year={2016},
pages={334-342},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0005752603340342},
isbn={978-989-758-173-1},
}

in EndNote Style

TY - CONF
JO - Proceedings of the 5th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM,
TI - A Pareto Front Approach for Feature Selection
SN - 978-989-758-173-1
AU - Grandchamp E.
AU - Abadi M.
AU - Alata O.
PY - 2016
SP - 334
EP - 342
DO - 10.5220/0005752603340342