Analysis of diversity methods for evolutionary multi-objective ensemble classifiers
Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review
Ensemble classifiers are strong and robust methods for classification and regression tasks. Considering the balance between runtime and classifier accuracy the learning problem becomes a multi-objective optimization problem. In this work, we propose an evolutionary multiobjective algorithm based on non-dominated sorting that balances runtime and accuracy properties of nearest neighbor classifier ensembles and decision tree ensembles. We identify relevant ensemble parameters with a significant impact on the accuracy and runtime. In the experimental part of this paper, we analyze the behavior on typical classification benchmark problems.
Original language | English |
---|---|
Title of host publication | Applications of Evolutionary Computation - 18th European Conference, EvoApplications 2015, Proceedings |
Editors | Giovanni Squillero, Antonio M. Mora |
Number of pages | 12 |
Publisher | Springer Verlag, |
Publication date | 1 Jan 2015 |
Pages | 567-578 |
ISBN (Electronic) | 9783319165486 |
DOIs | |
Publication status | Published - 1 Jan 2015 |
Externally published | Yes |
Event | 18th European Conference on the Applications of Evolutionary Computation, EvoApplications 2015 - Copenhagen, Denmark Duration: 8 Apr 2015 → 10 Apr 2015 |
Conference
Conference | 18th European Conference on the Applications of Evolutionary Computation, EvoApplications 2015 |
---|---|
Land | Denmark |
By | Copenhagen |
Periode | 08/04/2015 → 10/04/2015 |
Sponsor | Institute for Informatics and Digital Innovation, National Museum of Denmark, The World Federation on Soft Computing |
Series | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
---|---|
Volume | 9028 |
ISSN | 0302-9743 |
- Ensemble classification, Multi-objective optimization
Research areas
ID: 223196683