Journal of Algorithms and Computation
Open Access

Comparative Analysis of Metaheuristic Algorithms and Linear Regression in Predicting the CEO Selection

Document Type : Research Paper

Authors

hadi saeidi 1
Sara Hasanzadeh efati 1
Ghasem Elahi Shirvan 2
shaban Mohammadi 3

1 Department of Accounting, Shirvan Branch, Islamic Azad University, Shirvan, Iran

2 Department of Economy, Shirvan Branch, Islamic Azad University, Shirvan, Iran

3 Faculty of mathematical sciences, Shahrood university of technology, Shahrood, Semnan, Iran.

https://doi.org/10.22059/jac.2026.413114.1258
Abstract
The purpose of this study is to perform a comparative analysis of metaheuristic algorithms and linear regression in predicting the selection of CEO of companies listed on the Tehran Stock Exchange for 1050 data consisting of 15 companies for the years 2018 to 2024. The research data was analyzed by combining the particle swarm optimization algorithm and forbidden search algorithms, multilayer perceptron neural network, and gray wolf optimization, in order to predict the selection of CEO. the relationships between variables were examined using the regression method and the results obtained from the metaheuristic algorithms were compared. The results show that the variables of profitability, sales growth, return on assets, and managerial ability are important in all research algorithms in determining the type of CEO. Prediction based on the linear regression model is also possible. The results show the superiority of the neural network model.

Keywords

CEO Selection
Particle Swarm Optimization algorithm
Forbidden Search algorithm
Neural Network algorithm
Grey Wolf Optimization algorithm
Journal of Algorithms and Computation
Volume 58, Issue 1
July 2026
Pages 28-48

Home

Guide for Authors

Submit Manuscript

Reviewers

Contact Us