Abstract:
Evolutionary algorithms which model natural evolution processes have been successfully used for optimization. Theoretical explanations why and how the algorithms work have been less successful. In this paper evolutionary algorithms are considered as random search methods. The genetic operators mutation and recombination are evaluated according to the measure expected progress. Mutation is analyzed by Markov chains and probability theory. Recombination is investigated together with selection by methods of quantitative genetics. Furthermore it is shown that hillclimbing improves the efficiency of the search in ruggged landscapes. The theoretical results are used in two evolutionary algorithms- the parallel genetic algorithm PGA and the breeder genetic algorithm BGA. The PGA models natural evolution which self-organizes itself, the BGA models rational controlled evolution by a virtual breeder. The efficiency of both algorithms is shown with numerical examples. Results of a real world problem, the determination of binary sequences of low autocorrelation, are reported. 1
Citations
|
4827
|
Genetic Algorithms
– Goldberg
- 1989
|
|
465
|
Evolutionsstrategie: Optimierung Technischer Systeme nach Prinzipien der Biolgischen Evolution. Stuttgart-Bad Cannstatt: Verlag Frommann-Holzboog
– Rechenberg
- 1973
|
|
440
|
Uniform crossover in genetic algorithms
– Syswerda
- 1989
|
|
381
|
Numerical Optimization of Computer Models
– Schwefel
- 1981
|
|
277
|
A comparative analysis of selection schemes used in genetic algorithms, Foundations of Genetic Algorithms
– Goldberg, Deb
- 1991
|
|
271
|
Optimization of Control Parameters for Genetic Algorithms
– Grefenstette
- 1986
|
|
241
|
Predictive models for the breeder genetic algorithm
– Mühlenbein, Schlierkamp-Vose
- 1993
|
|
148
|
The parallel genetic algorithm as function optimizer
– Mühlenbein, Schomish, et al.
- 1991
|
|
99
|
Evolution in Time and Space – The Parallel Genetic Algorithm
– Mühlenbein
- 1991
|
|
76
|
Introduction to quantitative genetics
– Falconer, Mackay
- 1996
|
|
76
|
Evolution Algorithms in Combinatorial Optimization. Parallel Computing 7: 65–85
– Mühlenbein, Gorges-Schleuter, et al.
- 1988
|
|
56
|
Varying the probability of mutation in the genetic algorithm
– Fogarty
- 1989
|
|
50
|
Messy genetic algorithms revisited: studies in mixed size and scale
– Goldberg, Deb, et al.
- 1990
|
|
47
|
The mathematical theory of quantitative genetics
– Bulmer
- 1980
|
|
46
|
An introduction to population genetics theory. Harper and Row
– JF, Kimura
- 1970
|
|
34
|
Are genetic algorithms function optimizers
– DeJong
- 1992
|
|
29
|
Sieves for low autocorrelation binary sequences
– Golay
- 1977
|
|
28
|
Parallel genetic algorithm in combinatorial optimization
– Mühlenbein
- 1992
|
|
20
|
Global properties of evolution processes
– Bremermann, Rogson, et al.
- 1966
|
|
16
|
The merit factor of long low autocorrelation binary sequences
– Golay
- 1982
|
|
12
|
Binary sequences with a maximally flat amplitude spectrum
– Beenker, Claasen, et al.
- 1985
|
|
10
|
Parallel genetic algorithm, population dynamics and combinatorial optimization
– Muhlenbein
- 1989
|
|
8
|
Basic Concepts in Population, Quantitative and Evolutionary Genetics
– Crow
- 1986
|
|
6
|
Low autocorrelation binary sequences: exact enumeration and optimization by evolutionary strategies
– Groot, Würtz, et al.
- 1989
|
|
6
|
Optimization by Simulating Molecular Evolution
– Wang
- 1987
|
|
2
|
A theory of limits in artificial selection with many loci
– Robertson
- 1970
|
|
2
|
A study of control parameters effecting online performance of genetic algorithms for function optimization
– Schaffer, Caruana, et al.
- 1989
|
|
1
|
A new search for skewsymmetric binary se32 quences with optimal merit factors
– Golay
- 1989
|
|
