A Versatile Genetic Algorithm System in Ada: Implementation and Applications

Max Edwards




Abstract
Table of contents
<<< previous  Page 52 of 52

E.3 Time tabling problem genetic algorithm

See VAGAS/ttbl_ga/main.A

E.4 DeJong function genetic algorithm

See VAGAS/dejong_ga/main.A

E.5 Meta-level GA for TSP parameter-sets

See VAGAS/meta_tsp_ga/main.A

E.5.1 Table of fixed meta-parameters

The following parameters were used for all the meta-runs:

Bits per parameter

6

Maximum application fitness evaluations per run

2000

Maximum application generations per run

500

Crossover method

Simple

Mutation method

Simple

Selection method

Roulette

Scaling method

Simple Linear

E.5.2 Table of varying meta-parameters

The following parameters were used for the individual meta-runs:

Run

Pop. size

Num. gens.

Xover rate

Mut. rate

GA runs/eval.

Measure

1

128

128

0.75

0.05

1

off-line

2

256

512

0.75

0.05

1

off-line

3

256

64

0.75

0.05

3

off-line

4

128

128

0.75

0.05

5

off-line

5

128

128

0.75

0.05

8

off-line

6

128

250

0.85

0.02

8

off-line

7

128

250

0.85

0.02

16

on-line

8

256

127

0.9

0.02

16

on-line

9

128

128

0.85

0.02

16

on-line


Abstract
Table of contents
<<< previous  Page 52 of 52

Download source code     Download PDF of full report

© Copyright 1995 Max Edwards M.Eng.




  Google