## Decision Making In Genetic Algorithms: A Signal-To-Noise Perspective (1994)

Venue: | No 94004*, Illinois Genetic Algorithms Lab |

Citations: | 4 - 1 self |

### BibTeX

@TECHREPORT{Kargupta94decisionmaking,

author = {Hillol Kargupta and Hillol Kargupta and David E. Goldberg and David E. Goldberg},

title = {Decision Making In Genetic Algorithms: A Signal-To-Noise Perspective},

institution = {No 94004*, Illinois Genetic Algorithms Lab},

year = {1994}

}

### OpenURL

### Abstract

Signal detection in presence of noise is a decision problem, as addressed in the traditional signal processing literature. On the other hand, an arbitrary decision problem can also be posed in terms of signal and noise, corresponding to the inference to be tested. In this paper, we extend the previous efforts to analyze the decision making process in Genetic Algorithms(GAs) [Holland, 75] [Goldberg and Rudnick, 91] [Goldberg, Deb and Clark, 92] to a more general one for quantifying different aspects of difficulties that may lead GA towards a suboptimal solution. First we pose a decision problem in terms of multiple 2-armed bandits, which may be mutually dependent. This presents a theoretical framework for understanding stochastic search techniques in general. The deterministic and nondeterministic effects on the convolution kernel associated with the bandits are identified as signal and noise respectively. Next we discuss what these quantities mean in a GA. We also describe how they can...