University of Kerbala Discussing an M.A. Thesis entitled Estimation of Log-logistic Regression Parameters Using Genetic Algorithm with a Practical Application

Faculty of Administration and Economics / University of Kerbala has discussed the M.A. Thesis which is entitled “estimating log-logistic regression parameters using genetic algorithm with a practical application.”

The study , presented by Hussein Khalil Obaid Mukhalif, includes greatest possibility method, chi-square minimization method, and weighted least squares method.

The study aims at improving parameters of logistic regression model, making them less biased (closer to the real parameters), and building a logarithmic regression model and estimating its parameters for data related to known data.
The study concludes that the weighted least squares method is the best method among all the usual methods in estimating the model parameters.
The study recommends that conducting future studies dealing with dependence on usual estimation methods (MLE, WLSE, MCSE) as well as the methods improved by genetic algorithm (MLE.GA), (WLS.GA) and (MCSM.GA) through which the model parameters are estimated and compared between them. If there are outliers in the data that we will run the simulation experiment on.