Abstract
Multi-objective symbolic regression has the advantage that while the accuracy of the learned models is maximized, the complexity is automatically adapted and need not be specified a-priori. The result of the optimization is not a single solution anymore, but a whole Paretofront describing the trade-off between accuracy and complexity. In this contribution we study which complexity measures are most appropriately used in symbolic regression when performing multiobjective optimization with NSGA-II. Furthermore, we present a novel complexity measure that includes semantic information based on the function symbols occurring in the models and test its effects on several benchmark datasets. Results comparing multiple complexity measures are presented in terms of the achieved accuracy and model length to illustrate how the search direction of the algorithm is affected.
Original language | English |
---|---|
Title of host publication | Computer Aided Systems Theory – EUROCAST 2015 - 15th International Conference, Revised Selected Papers |
Editors | Franz Pichler, Roberto Moreno-Díaz, Alexis Quesada-Arencibia |
Publisher | Springer |
Pages | 409-416 |
Number of pages | 8 |
ISBN (Print) | 9783319273396 |
DOIs | |
Publication status | Published - 2015 |
Event | 15th International Conference on Computer Aided Systems Theory, Eurocast 2015 - Las Palmas, Gran Canaria, Spain Duration: 8 Feb 2015 → 13 Feb 2015 http://eurocast2015.fulp.ulpgc.es/ |
Publication series
Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
---|---|
Volume | 9520 |
ISSN (Print) | 0302-9743 |
ISSN (Electronic) | 1611-3349 |
Conference
Conference | 15th International Conference on Computer Aided Systems Theory, Eurocast 2015 |
---|---|
Country/Territory | Spain |
City | Las Palmas, Gran Canaria |
Period | 08.02.2015 → 13.02.2015 |
Internet address |
Keywords
- Complexity measures
- Genetic programming
- Multiobjective optimization
- NSGA-II
- Symbolic regression