DoE2Vec: Deep-learning Based Features for Exploratory Landscape Analysis

arXiv (Cornell University)

Abstract

We propose DoE2Vec, a variational autoencoder (VAE)-based methodology to learn optimization landscape characteristics for downstream meta-learning tasks, e.g., automated selection of optimization algorithms. Principally, using large training data sets generated with a random function generator, DoE2Vec self-learns an informative latent representation for any design of experiments (DoE). Unlike the classical exploratory landscape analysis (ELA) method, our approach does not require any feature engineering and is easily applicable for high dimensional search spaces. For validation, we inspect the quality of latent reconstructions and analyze the latent representations using different experiments. The latent representations not only show promising potentials in identifying similar (cheap-to-evaluate) surrogate functions, but also can significantly boost performances when being used complementary to the classical ELA features in classification tasks.

Authors 6

1. Bas van Stein Aachen

   Leiden University

   Affiliation as printed

   Leiden University Leiden , The Netherlands

2. Fu Xing Long
3. Moritz A. Frenzel

   BMW (Germany) · BMW Group (Germany)

   Affiliation as printed

   BMW Group Munich , Germany

4. Peter Krause
   Affiliation as printed

   divis intelligent solutions GmbH Dortmund , Germany

5. Markus Gitterle

   Munich University of Applied Sciences

   Affiliation as printed

   University of Applied Sciences Munich Munich , Germany

6. Thomas Bäck