Quality evaluations in optimisation processes are
frequently noisy. In particular evolutionary algorithms
have been shown to cope with such stochastic variations
better than other optimization algorithms. So far
mostly additive noise models have been assumed for the
analysis. However, we will argue in this paper that
this restriction must be relaxed for a large class of
applied optimization problems. We suggest systematic
noise as an alternative scenario, where the noise term
is added to the objective parameters or to
environmental parameters inside the fitness function.
We thoroughly analyse the sphere function with
systematic noise for the evolution strategy with global
intermediate recombination. The progress rate formula
and a measure for the efficiency of the evolutionary
progress lead to a recommended ratio between mu and
lambda. Furthermore, analysis of the dynamics
identifies limited regions of convergence dependent on
the normalized noise strength and the normalised
mutation strength. A residual localisation error
Rinfin can be quantified and a second mu to
lambda ratio is derived by minimising Rinfin.
%0 Journal Article
%1 beyer:2004:GPEM
%A Beyer, Hans-Georg
%A Olhofer, Markus
%A Sendhoff, Bernhard
%D 2004
%J Genetic Programming and Evolvable Machines
%K ES, analysis, evolution noisy optimisation, optimization performance robust strategies,
%N 4
%P 327--360
%R doi:10.1023/B:GENP.0000036020.79188.a0
%T On the Impact of Systematic Noise on the Evolutionary
Optimization Performance -- A Sphere Model Analysis
%V 5
%X Quality evaluations in optimisation processes are
frequently noisy. In particular evolutionary algorithms
have been shown to cope with such stochastic variations
better than other optimization algorithms. So far
mostly additive noise models have been assumed for the
analysis. However, we will argue in this paper that
this restriction must be relaxed for a large class of
applied optimization problems. We suggest systematic
noise as an alternative scenario, where the noise term
is added to the objective parameters or to
environmental parameters inside the fitness function.
We thoroughly analyse the sphere function with
systematic noise for the evolution strategy with global
intermediate recombination. The progress rate formula
and a measure for the efficiency of the evolutionary
progress lead to a recommended ratio between mu and
lambda. Furthermore, analysis of the dynamics
identifies limited regions of convergence dependent on
the normalized noise strength and the normalised
mutation strength. A residual localisation error
Rinfin can be quantified and a second mu to
lambda ratio is derived by minimising Rinfin.
@article{beyer:2004:GPEM,
abstract = {Quality evaluations in optimisation processes are
frequently noisy. In particular evolutionary algorithms
have been shown to cope with such stochastic variations
better than other optimization algorithms. So far
mostly additive noise models have been assumed for the
analysis. However, we will argue in this paper that
this restriction must be relaxed for a large class of
applied optimization problems. We suggest systematic
noise as an alternative scenario, where the noise term
is added to the objective parameters or to
environmental parameters inside the fitness function.
We thoroughly analyse the sphere function with
systematic noise for the evolution strategy with global
intermediate recombination. The progress rate formula
and a measure for the efficiency of the evolutionary
progress lead to a recommended ratio between [mu] and
[lambda]. Furthermore, analysis of the dynamics
identifies limited regions of convergence dependent on
the normalized noise strength and the normalised
mutation strength. A residual localisation error
R[infin] can be quantified and a second [mu] to
[lambda] ratio is derived by minimising R[infin].},
added-at = {2008-06-19T17:35:00.000+0200},
author = {Beyer, Hans-Georg and Olhofer, Markus and Sendhoff, Bernhard},
biburl = {https://www.bibsonomy.org/bibtex/27c3eb0bc6a6c3085faf86e5089a168fb/brazovayeye},
doi = {doi:10.1023/B:GENP.0000036020.79188.a0},
interhash = {35effa8d2f57cf8d1dde15ac844279e3},
intrahash = {7c3eb0bc6a6c3085faf86e5089a168fb},
issn = {1389-2576},
journal = {Genetic Programming and Evolvable Machines},
keywords = {ES, analysis, evolution noisy optimisation, optimization performance robust strategies,},
month = {December},
notes = {Article ID: 5272968},
number = 4,
pages = {327--360},
timestamp = {2008-06-19T17:36:33.000+0200},
title = {On the Impact of Systematic Noise on the Evolutionary
Optimization Performance -- {A} Sphere Model Analysis},
volume = 5,
year = 2004
}