We calculate the escape rates of different dynamical processes for the case of a one-dimensional symmetric double-well potential. In particular, we compare the escape rates of a Smoluchowski process, i.e., a corresponding overdamped Brownian motion dynamics in a metastable potential landscape, with the escape rates obtained for a biologically motivated model known as the Fisher-Eigen process. The main difference between the two models is that the dynamics of the Smoluchowski process is determined by local quantities, whereas the Fisher-Eigen process is based on a global coupling (nonlocal interaction). If considered in the context of numerical optimization algorithms, both processes can be interpreted as archetypes of physically or biologically inspired evolutionary strategies. In this sense, the results discussed in this work are utile in order to evaluate the efficiency of such strategies with regard to the problem of surmounting various barriers. We find that a combination of both scenarios, starting with the Fisher-Eigen strategy, provides a most effective evolutionary strategy.
[1]
Robert G. Reynolds,et al.
Evolutionary computation: Towards a new philosophy of machine intelligence
,
1997
.
[2]
Rainer Feistel,et al.
Evolution of complex systems : self-organization, entropy and development
,
1989
.
[3]
R. A. Fisher,et al.
The Genetical Theory of Natural Selection
,
1931
.
[4]
Ingo Rechenberg,et al.
Evolutionsstrategie : Optimierung technischer Systeme nach Prinzipien der biologischen Evolution
,
1973
.
[5]
John R. Koza,et al.
Genetic programming - on the programming of computers by means of natural selection
,
1993,
Complex adaptive systems.