CovSel: A new approach for ensemble selection applied to symbolic regression problems.
- Publikationstyp:
- Konferenzbeitrag
- Metadaten:
-
- DOI
- 10.1145/3205455
- ISBN-13
- 9781450356183
- Name of conference
- GECCO '18: Genetic and Evolutionary Computation Conference
- Online publication date
- 2018
- Datum der Veröffentlichung
- 2018
- Status
- Published
- Herausgeber
- ACM
- Herausgeber URL
- http://dx.doi.org/10.1145/3205455
- Datum der Datenerfassung
- 2023
- Titel
- Proceedings of the Genetic and Evolutionary Computation Conference
Datenquelle: Crossref
- Andere Metadatenquellen:
-
- Autoren
- Dominik Sobania
- Franz Rothlauf
- Editoren
- Hernán E Aguirre
- Keiki Takadama
- Zeitschrift
- GECCO
- Paginierung
- 529 - 536
- Datum der Veröffentlichung
- 2018
- Herausgeber
- ACM
- Herausgeber URL
- https://doi.org/10.1145/3205455
- Titel
- CovSel: A new approach for ensemble selection applied to symbolic regression problems.
Datenquelle: DBLP
- Abstract
- Ensemble methods combine the predictions of a set of models to reach a better prediction quality compared to a single model's prediction. The ensemble process consists of three steps: 1) the generation phase where the models are created, 2) the selection phase where a set of possible ensembles is composed and one is selected by a selection method, 3) the fusion phase where the individual models' predictions of the selected ensemble are combined to an ensemble's estimate. This paper proposes CovSel, a selection approach for regression problems that ranks ensembles based on the coverage of correctly estimated training points and selects the ensemble with the highest coverage to be used in the fusion phase. An ensemble covers a training point if at least one of its models produces a correct prediction for this training point. The more training points are covered this way, the higher is the ensemble's coverage. The selection of the "right'" ensemble has a large impact on the final prediction. Results for two symbolic regression problems show that CovSel improves the predictions for various state-of-the-art fusion methods for ensembles composed of indepentently evolved GP models and also beats approaches based on single GP models.
- Autoren
- Dominik Sobania
- Franz Rothlauf
- Autoren-URL
- https://dl.acm.org/authorize?N695221
- DOI
- 10.1145/3205455
- Editoren
- Hernán E Aguirre
- Keiki Takadama
- Conference finish date
- 2018
- Zeitschrift
- GECCO '18: Proceedings of the Genetic and Evolutionary Computation Conference
- Conference place
- Kyoto, Japan
- Name of conference
- Genetic and Evolutionary Computation Conference 2018
- Notes
- <!-- ACM DL Article: CovSel: A new approach for ensemble selection applied to symbolic regression problems --> <div class="acmdlitem" id="item3205570"><img src="//dl.acm.org/images/oa.gif" width="25" height="25" border="0" alt="ACM DL Author-ize service" style="vertical-align:middle"/><a href="https://dl.acm.org/authorize?N695221" title="CovSel: A new approach for ensemble selection applied to symbolic regression problems">CovSel: A new approach for ensemble selection applied to symbolic regression problems</a><div style="margin-left:25px"><a href="http://dl.acm.org/author_page.cfm?id=99659281071" >Dominik Sobania</a>, <a href="http://dl.acm.org/author_page.cfm?id=81100197550" >Franz Rothlauf</a><br />GECCO '18 Proceedings of the Genetic and Evolutionary Computation Conference, 2018</div></div> <!-- ACM DL Bibliometrics: CovSel: A new approach for ensemble selection applied to symbolic regression problems--> <div class="acmdlstat" id ="stats3205570"><iframe src="https://dl.acm.org/authorizestats?N695221" width="100%" height="30" scrolling="no" frameborder="0">frames are not supported</iframe></div>
- Paginierung
- 529 - 536
- Datum der Veröffentlichung
- 2018
- Herausgeber
- ACM
- Herausgeber URL
- https://dl.acm.org/authorize?N695221
- Datum der Datenerfassung
- 2019
- Conference start date
- 2018
- Titel
- CovSel: A new approach for ensemble selection applied to symbolic regression problems.
Datenquelle: Manual
- Beziehungen:
- Eigentum von