CMU Artificial Intelligence Repository
GAC: Simple GA in C
areas/genetic/ga/systems/gac/
This directory contains GAC, a simple GA (conceptually based on
Grefenstette's Genesis) written in C.
Baker's SUS selection algorithm is employed, n-point crossover is
maintained at 60%, and mutation is very low. Selection is based on
proportional fitness. This GA uses generations. It is also important
to note that this GA maximizes.
A note on crossover is in order. This version of GAC allows for
n-point crossover, where n is less than the length of an individual
(although there is no check for that). It is also possible to run
uniform crossover (see discussion below).
GAC will display run-time information as it executes. GAC also has
the ability to output this information into files. These statistics
include best behavior, online/ offline measurements, convergence, and
the number of reevaluations per generation. At this time the code is
commented out. The user can simply remove a few comment symbols to use
this facility. See run.c and geval.c for details.
There is no ranking, adaptive operators, etc. We intend to explore
these issues in future work.
See Also:
areas/genetic/ga/systems/gal/
Version: 9106.12
Requires: C
Copying: Property of the Department of the Navy.
Use, copying and distribution permitted.
CD-ROM: Prime Time Freeware for AI, Issue 1-1
Author(s): Bill Spears
Navy Center for Applied Research in AI
Naval Research Laboratory
Keywords:
Authors!Spears, C!Code, GAC, Genesis, Genetic Algorithms
References: ?
Last Web update on Mon Feb 13 10:22:47 1995
AI.Repository@cs.cmu.edu