A simple approach to lifetime learning in genetic programming-based symbolic regression
- Submitting institution
-
Birmingham City University
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 11Z_OP_D0009
- Type
- D - Journal article
- DOI
-
10.1162/EVCO_a_00111
- Title of journal
- Evolutionary computation
- Article number
- -
- First page
- 287
- Volume
- 22
- Issue
- 2
- ISSN
- 1063-6560
- Open access status
- Out of scope for open access requirements
- Month of publication
- -
- Year of publication
- 2014
- URL
-
https://www.mitpressjournals.org/doi/10.1162/EVCO_a_00111
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
-
- Research group(s)
-
-
- Citation count
- 19
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- Chameleon is the first method in Genetic Programming (GP) that - without additional computational cost - boosts quality of genetic search in GP with local search. Cost is contained because cheap models are innovatively preferred for local search, without subjectively penalising expensive models. Inducing efficiency thus is a new contribution.
Fear of computational cost has previously obstructed adopting local search in already expensive GP algorithms.
Chameleon gives theoretical guarantees over the limits of the costs of local search. The method also notes that existing theories of cost control subjectively penalise expensive models; instead, chameleon takes an organic approach.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -