Abstract: When developing and analyzing new hyperparameter optimization methods, it is vital to empirically evaluate and compare them on well-curated benchmark suites. In this work, we propose a new set of challenging and relevant benchmark problems motivated by desirable properties and requirements for such benchmarks. Our new surrogate-based benchmark collection consists of 14 scenarios that in total constitute over 700 multi-fidelity hyperparameter optimization problems, which all enable multi-objective hyperparameter optimization. Furthermore, we empirically compare surrogate-based benchmarks to the more widely-used tabular benchmarks, and demonstrate that the latter may produce unfaithful results regarding the performance ranking of HPO methods. We examine and compare our benchmark collection with respect to defined requirements and propose a single-objective as well as a multi-objective benchmark suite on which we compare 7 single-objective and 7 multi-objective optimizers in a benchmark experiment. Our software is available at https://github.com/slds-lmu/yahpo_gym.
Keywords: HPO, optimization, benchmarking, multi-fidelity, multi-objective, NAS
One-sentence Summary: We propose a new benchmark collection for multi-fidelity and multi-objective HPO.
Track: Special track for systems, benchmarks and challenges
Reproducibility Checklist: Yes
Broader Impact Statement: Yes
Paper Availability And License: Yes
Code Of Conduct: Yes
Reviewers: Florian Pfisterer is already reviewer. Bernd Bischl is already senior area chair.
Main Paper And Supplementary Material: pdf
Steps For Environmental Footprint Reduction During Development: Instead of tuning and fitting surrogate models for each individual benchmark problem, we fit our surrogates on whole benchmark scenarios (multiple instances and multiple targets).
CPU Hours: 3349
GPU Hours: 1080
TPU Hours: 0
Evaluation Metrics: Yes
Estimated CO2e Footprint: 381