Beta
220870

SOLVING TWO-CLASS CLASSIFICATION PROBLEM USING MEMETIC PROGRAMMING

Article

Last updated: 23 Jan 2023

Subjects

-

Tags

Mathematics and Computer Science

Abstract

The main target of the two-class classification problem is to design a classifier that discriminates between two objects from a seen dataset, then use this classifier to predict the object's class for unseen instances. Different methods have been used to solve the two-class classification problem, such as Genetic algorithm (GA) and Genetic Programming (GP). However, there is still a need to design new methods that can overcome some limitations in evolutionary algorithms, e.g., the high disruption of the breading operations; mutation and crossover. Recently, the Memetic Programming (MP) algorithm was proposed as an improvement to the GP algorithm. In this paper, we adapt the MP algorithm to produce a new classifier algorithm called the Memetic Programming Classifier (MPC) algorithm to solve the two-class classification problem. The performance of MPC is validated through different datasets from the UCI database and the accuracy is compared along with different methods. As a result, the proposed MPC algorithm shows a competitive performance compared with 179 classifiers in the literature.

DOI

10.21608/aunj.2019.220870

Keywords

Classifier, Classification Problem, Genetic Programming, Local Search, Memetic Programming

Volume

48

Article Issue

1

Related Issue

31480

Issue Date

2019-06-01

Receive Date

2022-02-21

Publish Date

2019-06-01

Page Start

58

Page End

71

Print ISSN

2812-5029

Online ISSN

2812-5037

Link

https://aunj.journals.ekb.eg/article_220870.html

Detail API

https://aunj.journals.ekb.eg/service?article_code=220870

Order

220,870

Type

Novel Research Articles

Type Code

2,242

Publication Type

Journal

Publication Title

Assiut University Journal of Multidisciplinary Scientific Research

Publication Link

https://aunj.journals.ekb.eg/

MainTitle

-

Details

Type

Article

Created At

23 Jan 2023