Benchmarking GNN-CMA-ES on the BBOB noiseless testbed
| Autor: | Olivier Fercoq, Louis Faury, Clément Calauzènes |
|---|---|
| Přispěvatelé: | Criteo [Paris], Laboratoire Traitement et Communication de l'Information (LTCI), Institut Mines-Télécom [Paris] (IMT)-Télécom Paris, Fercoq, Olivier |
| Jazyk: | angličtina |
| Rok vydání: | 2019 |
| Předmět: |
010308 nuclear & particles physics
Computer science Testbed Contrast (statistics) [MATH.MATH-OC] Mathematics [math]/Optimization and Control [math.OC] Multivariate normal distribution 02 engineering and technology Extension (predicate logic) Benchmarking 01 natural sciences 0103 physical sciences 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing [MATH.MATH-OC]Mathematics [math]/Optimization and Control [math.OC] CMA-ES Algorithm |
| Zdroj: | the Genetic and Evolutionary Computation Conference Companion the Genetic and Evolutionary Computation Conference Companion, Jul 2019, Prague, Czech Republic. pp.1928-1936 GECCO (Companion) |
| Popis: | International audience; Popular machine learning estimators involve regularization parameters that can be challenging to tune, and standard strategies rely on grid search for this task. In this paper, we revisit the techniques of approximating the regularization path up to predefined tolerance $\epsilon$ in a unified framework and show that its complexity is $O(1/\sqrt[d]{\epsilon})$ for uniformly convex loss of order $d \geq 2$ and $O(1/\sqrt{\epsilon})$ for Generalized Self-Concordant functions. This framework encompasses least-squares but also logistic regression, a case that as far as we know was not handled as precisely in previous works. We leverage our technique to provide refined bounds on the validation error as well as a practical algorithm for hyperparameter tuning. The latter has global convergence guarantee when targeting a prescribed accuracy on the validation set. Last but not least, our approach helps relieving the practitioner from the (often neglected) task of selecting a stopping criterion when optimizing over the training set: our method automatically calibrates this criterion based on the targeted accuracy on the validation set. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|