| Summary: | Ant Colony Optimization (ACO) has been used to solve Support Vector Machine (SVM) model selection problem.ACO originally deals with discrete optimization problem. In applying ACO for optimizing SVM parameters which are continuous variables, there is a need to discretize the continuously value into discrete values.This discretize process would result in loss of some information and hence affect the classification accuracy.In order to enhance SVM performance and solving the discretization problem, this study proposes two algorithms to optimize SVM parameters using Continuous ACO (ACOR) and Incremental Continuous Ant Colony Optimization (IACOR) without the need to discretize continuous value for SVM parameters.Eight datasets from UCI were used to evaluate the credibility of the proposed integrated algorithm in terms of classification accuracy and size of features subset.Promising results were obtained when compared to grid search technique, GA with feature chromosome-SVM, PSO-SVM, and GA-SVM. Results have also shown that IACOR-SVM is better than ACOR-SVM in terms of classification accuracy.
|
|---|