Q(D)O-ES: Population-based Quality (Diversity) Optimisation for Post Hoc Ensemble Selection in AutoML

arXiv (Cornell University)

Abstract

Automated machine learning (AutoML) systems commonly ensemble models post hoc to improve predictive performance, typically via greedy ensemble selection (GES). However, we believe that GES may not always be optimal, as it performs a simple deterministic greedy search. In this work, we introduce two novel population-based ensemble selection methods, QO-ES and QDO-ES, and compare them to GES. While QO-ES optimises solely for predictive performance, QDO-ES also considers the diversity of ensembles within the population, maintaining a diverse set of well-performing ensembles during optimisation based on ideas of quality diversity optimisation. The methods are evaluated using 71 classification datasets from the AutoML benchmark, demonstrating that QO-ES and QDO-ES often outrank GES, albeit only statistically significant on validation data. Our results further suggest that diversity can be beneficial for post hoc ensembling but also increases the risk of overfitting.

Authors 6

1. Lennart Purucker

   University of Siegen

   Affiliation as printed

   University of Siegen ,

2. Lennart Schneider

   Munich Center for Machine Learning · Ludwig-Maximilians-Universität München

   Affiliation as printed

   LMU Munich ,

   Munich Center for Machine Learning (MCML) ,

3. Marie Anastacio Aachen

   RWTH Aachen University

   Affiliation as printed

   RWTH Aachen ,

4. Joeran Beel

   University of Siegen

   Affiliation as printed

   University of Siegen ,

5. Bernd Bischl

   Munich Center for Machine Learning · Ludwig-Maximilians-Universität München

   Affiliation as printed

   LMU Munich ,

   Munich Center for Machine Learning (MCML) ,

6. Holger H. Hoos