Random Rule Sets – Combining Random Covering with the Random Subspace Method
Tony Lindgren · International Journal of Machine Learning and Computing · 2018
Ensembles of classifiers have been proven to be among the best methods to create highly accurate prediction models.In this study, we combine the random coverage method, which facilitates additional diversity when inducing rules using the covering algorithm, with the random subspace selection method, which has been used successfully by the random forest algorithm.We compare three different covering methods with the random forest algorithm: (1) using random subspace selection and random coverage, (2) using bagging and random subspace selection, and (3) using bagging, random subspace selection, and random coverage.The results show that all three covering algorithms perform better than the random forest algorithm.The covering algorithm using random subspace selection and random coverage performs best among all methods.The results are not significant according to adjusted p values but they are according to unadjusted p values, indicating that the novel method introduced in this study warrants further attention.