Bayesian Hyperparameter Optimization for Ensemble Learning

TL;DR

This paper introduces a Bayesian hyperparameter optimization method for ensemble learning that optimizes ensemble configurations considering interactions among models, leading to improved performance over single models and standard ensemble methods.

Contribution

The paper presents a novel Bayesian optimization approach that jointly optimizes ensemble hyperparameters and configurations, including a greedy ensemble reconstruction strategy.

Findings

Our method outperforms the best single models.

It surpasses standard Bayesian optimization in ensemble performance.

Ensemble reconstruction enhances overall results.

Abstract

In this paper, we bridge the gap between hyperparameter optimization and ensemble learning by performing Bayesian optimization of an ensemble with regards to its hyperparameters. Our method consists in building a fixed-size ensemble, optimizing the configuration of one classifier of the ensemble at each iteration of the hyperparameter optimization algorithm, taking into consideration the interaction with the other models when evaluating potential performances. We also consider the case where the ensemble is to be reconstructed at the end of the hyperparameter optimization phase, through a greedy selection over the pool of models generated during the optimization. We study the performance of our proposed method on three different hyperparameter spaces, showing that our approach is better than both the best single model and a greedy ensemble construction over the models produced by a standard Bayesian optimization.