A Hybrid of Artificial Bee Colony, Genetic Algorithm, and Neural Network for Diabetic Mellitus Diagnosing

Rashid, Tarik and Abdullah, Saman (2018) A Hybrid of Artificial Bee Colony, Genetic Algorithm, and Neural Network for Diabetic Mellitus Diagnosing. ARO-The Scientific Journal of Koya University, 6 (1). pp. 55-64. ISSN 24109355

[img]
Preview
Text (PDF file)
ARO.10368-VOL6.No1.2018.ISSUE10-PP55-64.pdf - Published Version
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (3MB) | Preview
Official URL: http://dx.doi.org/10.14500/aro.10368

Abstract

Researchers, widely have introduced the Artificial Bee Colony (ABC) as an optimization algorithm to deal with classification, and prediction problems. ABC has been combined with different Artificial Intelligent (AI) techniques to obtain optimum performance indicators. This work introduces a hybrid of ABC, Genetic Algorithm (GA), and Back Propagation Neural Network (BPNN) in the application of classifying, and diagnosing Diabetic Mellitus (DM). The optimized algorithm is combined with a mutation technique of Genetic Algorithm (GA) to obtain the optimum set of training weights for a BPNN. The idea is to prove that weights’ initial index in their initialized set has an impact on the performance rate. Experiments are conducted in three different cases; standard BPNN alone, BPNN trained with ABC, and BPNN trained with the mutation based ABC. The work tests all three cases of optimization on two different datasets (Primary dataset, and Secondary dataset) of diabetic mellitus (DM). The primary dataset is built by this work through collecting 31 features of 501 DM patients in local hospitals. The secondary dataset is the Pima dataset. Results show that the BPNN trained with the mutation based ABC can produce better local solutions than the standard BPNN and BPNN trained in combination with ABC.

Item Type: Article
Uncontrolled Keywords: Artificial Bee Colony, Artificial Neural Networks, Diabetic Mellitus, Evolutionary Algorithms.
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Divisions: ARO-The Scientific Journal of Koya University > VOL 6, NO 1 (2018)
Depositing User: Dr Salah Ismaeel Yahya
Date Deposited: 24 Oct 2018 07:39
Last Modified: 30 Mar 2020 22:49
URI: http://eprints.koyauniversity.org/id/eprint/167

Actions (login required)

View Item View Item