Conjugate Schema and Basis Representation of Crossover and Mutation Operators
- Creators
- Kazadi, Sanza T.
Abstract
In genetic search algorithms and optimization routines, the representation of the mutation and crossover operators are typically defaulted to the canonical basis. We show that this can be influential in the usefulness of the search algorithm. We then pose the question of how to find a basis for which the search algorithm is most useful. The conjugate schema is introduced as a general mathematical construct and is shown to separate a function into smaller dimensional functions whose sum is the original function. It is shown that conjugate schema, when used on a test suite of functions, improves the performance of the search algorithm on 10 out of 12 of these functions. Finally, a rigorous but abbreviated mathematical derivation is given in the appendices.
Additional Information
© 1998 Massachusetts Institute of Technology.Attached Files
Published - evco.1998.6.2.129.pdf
Files
Name | Size | Download all |
---|---|---|
md5:3ee6d94e029f255b7c9b5122a6fa5fc2
|
2.0 MB | Preview Download |
Additional details
- Eprint ID
- 72038
- Resolver ID
- CaltechAUTHORS:20161115-173034535
- Created
-
2016-11-17Created from EPrint's datestamp field
- Updated
-
2021-11-11Created from EPrint's last_modified field