loading
Papers Papers/2022 Papers Papers/2022

Research.Publish.Connect.

Paper

Paper Unlock

Authors: Ruba Abu Khurma 1 ; Pedro A. Castillo 2 ; Ahmad Sharieh 1 and Ibrahim Aljarah 1

Affiliations: 1 King Abdullah II School for Information Technology, The University of Jordan, Amman, Jordan ; 2 Department of Computer Architecture and Computer Technology, ETSIIT and CITIC, University of Granada, Granada, Spain

Keyword(s): Moth Flame Optimization, MFO, Feature Selection, Classification, Flames Number, Optimization.

Abstract: In this paper, a new feature selection (FS) approach is proposed based on the Moth Flame Optimization (MFO) algorithm with time-varying flames number strategies. FS is a data preprocessing technique that is applied to minimize the number of features in a data set to enhance the performance of the learning algorithm (e.g classifier) and reduce the learning time. Finding the best feature subset is a challenging search process that requires exponential running time if the complete search space is generated. Meta-heuristics algorithms are promising alternative solutions that have proven their performance in finding approximated optimal solutions within a reasonable time. The MFO algorithm is a recently developed Swarm Intelligence (SI) algorithm that has demonstrated effective performance in solving various optimization problems. This is due to its spiral update strategy that enhances the convergence trends of the algorithm. The number of flames is an important parameter in the MFO algor ithm that controls the balance between the exploration and exploitation phases during the optimization process. In the standard MFO, the number of flames linearly decreases throughout the iterations. This paper proposes different time-varying strategies to update the number of flames and analyzes their impact on the performance of MFO when used to solve the FS problem. Seventeen medical benchmark data sets were used to evaluate the performance of the proposed approach. The proposed approach is compared with other well-regarded meta-heuristics and the results show promising performance in tackling the FS problem. (More)

CC BY-NC-ND 4.0

Sign In Guest: Register as new SciTePress user now for free.

Sign In SciTePress user: please login.

PDF ImageMy Papers

You are not signed in, therefore limits apply to your IP address 18.117.145.67

In the current month:
Recent papers: 100 available of 100 total
2+ years older papers: 200 available of 200 total

Paper citation in several formats:
Khurma, R.; Castillo, P.; Sharieh, A. and Aljarah, I. (2020). Feature Selection using Binary Moth Flame Optimization with Time Varying Flames Strategies. In Proceedings of the 12th International Joint Conference on Computational Intelligence (IJCCI 2020) - ECTA; ISBN 978-989-758-475-6; ISSN 2184-3236, SciTePress, pages 17-27. DOI: 10.5220/0010021700170027

@conference{ecta20,
author={Ruba Abu Khurma. and Pedro A. Castillo. and Ahmad Sharieh. and Ibrahim Aljarah.},
title={Feature Selection using Binary Moth Flame Optimization with Time Varying Flames Strategies},
booktitle={Proceedings of the 12th International Joint Conference on Computational Intelligence (IJCCI 2020) - ECTA},
year={2020},
pages={17-27},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0010021700170027},
isbn={978-989-758-475-6},
issn={2184-3236},
}

TY - CONF

JO - Proceedings of the 12th International Joint Conference on Computational Intelligence (IJCCI 2020) - ECTA
TI - Feature Selection using Binary Moth Flame Optimization with Time Varying Flames Strategies
SN - 978-989-758-475-6
IS - 2184-3236
AU - Khurma, R.
AU - Castillo, P.
AU - Sharieh, A.
AU - Aljarah, I.
PY - 2020
SP - 17
EP - 27
DO - 10.5220/0010021700170027
PB - SciTePress