Handling High-Dimensional Regression Problems by Means of an Efficient Multi-Objective Evolutionary Algorithm

Linguistic fuzzy modeling in high dimensional regression problems is a challenging topic since conventional linguistic fuzzy rule-based systems suffer from exponential rule explosion when the number of variables and/or data examples becomes high. A good way to face this problem is by searching for a good and simple global structure within the same process, in order to consider the relationships among the different components defining the final linguistic model. In this contribution, we propose an effective multi-objective evolutionary algorithm that based on the data base learning a priori (involved variables, granularities and slight uniform displacements of the fuzzy partitions) allows a fast derivation of simple and quite accurate linguistic models, making use of some effective mechanisms in order to ensure a fast convergence. The good results obtained in several large-scale regression problems demonstrate the effectiveness of the proposed approach.