توسعه الگوریتم بهینه سازی کلونی مورچگان (ACO) بر اساس تجزیه و تحلیل آماری و تست فرضیه برای انتخاب متغیر

Obtaining reliable models from experimental data is a point of deep interest in all areas of research. Since the quality of the model depends on the number of selected variables, it is important to develop methods that identify the best ones. This work proposes a method of variable selection based on the Ant Colony Optimization (ACO) algorithm. Using data from a Saccharomyces cerevisiae fermentation, several criteria for trail update and model comparison were implemented and the obtained models were compared. The use of the length of the confidence interval produced the best results, finding the optimal model more frequently.