Selecting the selector: Comparison of update rules for discrete global optimization

Autor:	James Theiler, Beate G. Zimmer
Rok vydání:	2017
Předmět:	Statistics::Theory Computer science Bayesian probability 02 engineering and technology Machine learning computer.software_genre 01 natural sciences 010104 statistics & probability Statistics::Methodology Point (geometry) 0101 mathematics Global optimization business.industry Function (mathematics) 021001 nanoscience & nanotechnology Computer Science Applications Term (time) Noise Level of measurement Artificial intelligence 0210 nano-technology business Focus (optics) Algorithm computer Analysis Information Systems
Zdroj:	Statistical Analysis and Data Mining: The ASA Data Science Journal. 10:211-229
ISSN:	1932-1872 1932-1864
DOI:	10.1002/sam.11343
Popis:	We compare some well-known Bayesian global optimization methods in four distinct regimes, corresponding to high and low levels of measurement noise and to high and low levels of “quenched noise” (which term we use to describe the roughness of the function we are trying to optimize). We isolate the two stages of this optimization in terms of a “regressor,” which fits a model to the data measured so far, and a “selector,” which identifies the next point to be measured. The focus of this paper is to investigate the choice of selector when the regressor is well matched to the data.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::f69cf8ee290409b3ff9009e39be983c6 https://doi.org/10.1002/sam.11343 Zobrazit plný text záznamu Plný text ve formátu PDF Plný text ve formátu HTML
Nepřihlášeným uživatelům se plný text nezobrazuje	K zobrazení výsledku je třeba se přihlásit.