A PageRank-Based Method to Extract Fuzzy Expressions as Features in Supervised Classification Problems

摘要

This work presents a new ranking method inspired on PageRank to reduce the dimensionality of the feature space in supervised classification problems. More precisely, as it relies on a weighted directed graph, it is ultimately inspired on TextRank, a PageRank based method that adds weights to the edges to express the strength of the connections between nodes. The method is based on dividing each original feature used to describe the training set into a set of fuzzy predicates and then ranking all of them by their ability to differentiate among classes in the light of this training set. The fuzzy predicates with the best scores can be then used as new features, replacing the original ones. The novelty of the proposal relies on being an approach halfway between feature selection and feature extraction approaches, being able to improve the discrimination ability of the original features but preserving the interpret ability of the new features in the sense that they are fuzzy expressions. Preliminary results supports the suitability of the proposal.