Examples of inconsistency in optimization by expected improvement

We consider the 1D Expected Improvement optimization based on Gaussian processes having spectral densities converging to zero faster than exponentially. We give examples of problems where the optimization trajectory is not dense in the design space. In particular, we prove that for Gaussian kernels there exist smooth objective functions for which the optimization does not converge on the optimum. Copyright Springer Science+Business Media, LLC. 2013

MoreLess

Year of publication:	2013
Authors:	Yarotsky, Dmitry
Published in:	Journal of Global Optimization. - Springer. - Vol. 56.2013, 4, p. 1773-1790
Publisher:	Springer
Subject:	Bayesian optimization \| Expected improvement \| Gaussian covariance function \| Analytic kernel

Online Resource

Check full text access |

More access options

Check Google Scholar

In German libraries (KVK)

I need help

More details

Extent:	text/html
Type of publication:	Article
Source:	RePEc - Research Papers in Economics

Persistent link: https://www.econbiz.de/10010896418