Národní úložiště šedé literatury Nalezeno 1 záznamů.  Hledání trvalo 0.02 vteřin. 
Platforms for bechmarking optimization algorithms for continuous problems
Visingr, Patrik ; Shehadeh, Mhd Ali (oponent) ; Kůdela, Jakub (vedoucí práce)
This thesis explores benchmarking platforms for comparing algorithms using black-box methods. It begins with an overview of optimization algorithms and detailed examination of three prominent benchmarking platforms: COCO, IOHprofiler, and Nevergrad. These platforms are instrumental in evaluating optimization algorithms by providing standardized frameworks, comprehensive performance metrics, and ensuring reproducibility of experiments. This collective approach aids in identifying the best algorithms for specific applications and optimizing their performance through parameter tuning. It also highlights the advantages and disadvantages of each platform and their cooperative capabilities.

Chcete být upozorněni, pokud se objeví nové záznamy odpovídající tomuto dotazu?
Přihlásit se k odběru RSS.