Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

PHS: A Toolbox for Parallel Hyperparameter Search

: Habelitz, Peter Michael; Keuper, Janis

Volltext ()

Online im WWW, 2020, arXiv:2002.11429, 5 S.
Bundesministerium für Bildung und Forschung BMBF (Deutschland)
Deep Topology Learning
Bericht, Elektronische Publikation
Fraunhofer ITWM ()

We introduce an open source python framework named PHS - Parallel Hyperparameter Search to enable hyperparameter optimization on numerous compute instances of any arbitrary python function. This is achieved with minimal modifications inside the target function. Possible applications appear in expensive to evaluate numerical computations which strongly depend on hyperparameters such as machine learning. Bayesian optimization is chosen as a sample efficient method to propose the next query set of parameters.