A comparison between genetic algorithm and weighted least squares for estimating a multiple response logistic regression model using simulation

##plugins.themes.academic_pro.article.main##

root root

Abstract

The logistic regression model is one of the most important regression analysis models in the case of metadata. Where this research included a comparison between the genetic algorithm method with the weighted least squares method, where a simulation experiment was conducted for different sample sizes (small, medium, large) (125,75,25). For three cases of default values ​​that were set by the researcher for the values ​​of the model parameters (B2, B1, B0) using the Monte-Carlo method for the purpose of arriving at the optimal method for estimating a multi-level logistic regression model, and the results showed that the genetic algorithm (GA) method It is better and more efficient than the weighted least squares (WLS) method at small sample sizes (25).

##plugins.themes.academic_pro.article.details##

How to Cite
root, root. (2022). A comparison between genetic algorithm and weighted least squares for estimating a multiple response logistic regression model using simulation. Warith Scientific Journal, 4(12), 255-267. https://doi.org/10.57026/wsj.v4i12.89