Abstract
We propose a new boosting algorithm based on a linear programming formulation. Our algorithm can take advantage of the sparsity of the solution of the underlying optimization problem. In preliminary experiments, our algorithm outperforms a state-of-the-art LP solver and LPBoost especially when the solution is given by a small set of relevant hypotheses and support vectors.
| Original language | English |
|---|---|
| Title of host publication | Discovery Science - 12th International Conference, DS 2009, Proceedings |
| Pages | 401-408 |
| Number of pages | 8 |
| DOIs | |
| Publication status | Published - Nov 16 2009 |
| Event | 12th International Conference on Discovery Science, DS 2009 - Porto, Portugal Duration: Oct 3 2009 → Oct 5 2009 |
Publication series
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Volume | 5808 LNAI |
| ISSN (Print) | 0302-9743 |
| ISSN (Electronic) | 1611-3349 |
Other
| Other | 12th International Conference on Discovery Science, DS 2009 |
|---|---|
| Country | Portugal |
| City | Porto |
| Period | 10/3/09 → 10/5/09 |
Fingerprint
All Science Journal Classification (ASJC) codes
- Theoretical Computer Science
- Computer Science(all)
Cite this
Linear programming boosting by column and row generation. / Hatano, Kohei; Takimoto, Eiji.
Discovery Science - 12th International Conference, DS 2009, Proceedings. 2009. p. 401-408 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5808 LNAI).Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
}
TY - GEN
T1 - Linear programming boosting by column and row generation
AU - Hatano, Kohei
AU - Takimoto, Eiji
PY - 2009/11/16
Y1 - 2009/11/16
N2 - We propose a new boosting algorithm based on a linear programming formulation. Our algorithm can take advantage of the sparsity of the solution of the underlying optimization problem. In preliminary experiments, our algorithm outperforms a state-of-the-art LP solver and LPBoost especially when the solution is given by a small set of relevant hypotheses and support vectors.
AB - We propose a new boosting algorithm based on a linear programming formulation. Our algorithm can take advantage of the sparsity of the solution of the underlying optimization problem. In preliminary experiments, our algorithm outperforms a state-of-the-art LP solver and LPBoost especially when the solution is given by a small set of relevant hypotheses and support vectors.
UR - http://www.scopus.com/inward/record.url?scp=71049179485&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=71049179485&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-04747-3_32
DO - 10.1007/978-3-642-04747-3_32
M3 - Conference contribution
AN - SCOPUS:71049179485
SN - 3642047467
SN - 9783642047466
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 401
EP - 408
BT - Discovery Science - 12th International Conference, DS 2009, Proceedings
ER -