Mean square rates of convergence in the continuous time simulated annealing algorithm on Rd

Authors:
Larry Goldstein
Affiliations:
Department of Mathematics, University of Southern California, Los Angeles, California 90089-1113 U.S.A.
Venue:
Advances in Applied Mathematics
Year:
1988

Citing 0
Cited 1

The Biological Concept of Neoteny in Evolutionary Color Image Segmentation - Simple Experiments in Simple Non-memetic Genetic Algorithms

Proceedings of the EvoWorkshops on Applications of Evolutionary Computing

Quantified Score

Hi-index	0.00

Visualization

Abstract

To locate a global minimum value of a function V: R^d - R one may employ the continuous time simulated annealing algorithm dX"t = -@? V(X"t) dt + @s"tdW"t; t =0, E@?X"0@?^2 " "~@s"t = 0 and @s"t = @/c'log tfor t sufficiently large. If @s"t = @/clog(t + 2)and V is a function with global minimum at @q @e R^d and any number of local minima then under suitable conditions there exists a positive constant c such that E@?@/log t(X"t - @q)@?^2 = c, for all t sufficiently large.