Classifier Pool Generation based on a Two-level Diversity Approach

TL;DR

This paper introduces a novel classifier pool generation method that uses a two-level diversity approach based on data complexity and classifier decisions, leading to improved accuracy in ensemble methods.

Contribution

It proposes a new diversity-guided pool generation technique using complexity measures and evolutionary algorithms, enhancing classifier ensemble performance.

Findings

Significant accuracy improvements in 69.4% of experiments

Effective selection of complexity measures with high variability

Enhanced performance of dynamic classifier and ensemble selection methods

Abstract

This paper describes a classifier pool generation method guided by the diversity estimated on the data complexity and classifier decisions. First, the behavior of complexity measures is assessed by considering several subsamples of the dataset. The complexity measures with high variability across the subsamples are selected for posterior pool adaptation, where an evolutionary algorithm optimizes diversity in both complexity and decision spaces. A robust experimental protocol with 28 datasets and 20 replications is used to evaluate the proposed method. Results show significant accuracy improvements in 69.4% of the experiments when Dynamic Classifier Selection and Dynamic Ensemble Selection methods are applied.