Details
| Original language | English |
|---|---|
| Pages (from-to) | 41-58 |
| Number of pages | 18 |
| Journal | Artificial intelligence |
| Volume | 237 |
| Early online date | 8 Apr 2016 |
| Publication status | Published - Apr 2016 |
| Externally published | Yes |
Abstract
The task of algorithm selection involves choosing an algorithm from a set of algorithms on a per-instance basis in order to exploit the varying performance of algorithms over a set of instances. The algorithm selection problem is attracting increasing attention from researchers and practitioners in AI. Years of fruitful applications in a number of domains have resulted in a large amount of data, but the community lacks a standard format or repository for this data. This situation makes it difficult to share and compare different approaches effectively, as is done in other, more established fields. It also unnecessarily hinders new researchers who want to work in this area. To address this problem, we introduce a standardized format for representing algorithm selection scenarios and a repository that contains a growing number of data sets from the literature. Our format has been designed to be able to express a wide variety of different scenarios. To demonstrate the breadth and power of our platform, we describe a study that builds and evaluates algorithm selection models through a common interface. The results display the potential of algorithm selection to achieve significant performance improvements across a broad range of problems and algorithms.
Keywords
- Algorithm selection, Empirical performance estimation, Machine learning
ASJC Scopus subject areas
- Arts and Humanities(all)
- Language and Linguistics
- Social Sciences(all)
- Linguistics and Language
- Computer Science(all)
- Artificial Intelligence
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
In: Artificial intelligence, Vol. 237, 04.2016, p. 41-58.
Research output: Contribution to journal › Article › Research › peer review
}
TY - JOUR
T1 - ASlib: A benchmark library for algorithm selection
AU - Bischl, Bernd
AU - Kerschke, Pascal
AU - Kotthoff, Lars
AU - Lindauer, Marius
AU - Malitsky, Yuri
AU - Fréchette, Alexandre
AU - Hoos, Holger
AU - Hutter, Frank
AU - Leyton-Brown, Kevin
AU - Tierney, Kevin
AU - Vanschoren, Joaquin
N1 - Funding information: FH and ML are supported by the DFG (German Research Foundation) under Emmy Noether grant HU 1900/2-1 . KLB, AF and LK were supported by an NSERC E.W.R. Steacie Fellowship; in addition, all of these, along with HH, were supported under the NSERC Discovery Grant Program. Part of this research was supported by a Microsoft Azure for Research grant.
PY - 2016/4
Y1 - 2016/4
N2 - The task of algorithm selection involves choosing an algorithm from a set of algorithms on a per-instance basis in order to exploit the varying performance of algorithms over a set of instances. The algorithm selection problem is attracting increasing attention from researchers and practitioners in AI. Years of fruitful applications in a number of domains have resulted in a large amount of data, but the community lacks a standard format or repository for this data. This situation makes it difficult to share and compare different approaches effectively, as is done in other, more established fields. It also unnecessarily hinders new researchers who want to work in this area. To address this problem, we introduce a standardized format for representing algorithm selection scenarios and a repository that contains a growing number of data sets from the literature. Our format has been designed to be able to express a wide variety of different scenarios. To demonstrate the breadth and power of our platform, we describe a study that builds and evaluates algorithm selection models through a common interface. The results display the potential of algorithm selection to achieve significant performance improvements across a broad range of problems and algorithms.
AB - The task of algorithm selection involves choosing an algorithm from a set of algorithms on a per-instance basis in order to exploit the varying performance of algorithms over a set of instances. The algorithm selection problem is attracting increasing attention from researchers and practitioners in AI. Years of fruitful applications in a number of domains have resulted in a large amount of data, but the community lacks a standard format or repository for this data. This situation makes it difficult to share and compare different approaches effectively, as is done in other, more established fields. It also unnecessarily hinders new researchers who want to work in this area. To address this problem, we introduce a standardized format for representing algorithm selection scenarios and a repository that contains a growing number of data sets from the literature. Our format has been designed to be able to express a wide variety of different scenarios. To demonstrate the breadth and power of our platform, we describe a study that builds and evaluates algorithm selection models through a common interface. The results display the potential of algorithm selection to achieve significant performance improvements across a broad range of problems and algorithms.
KW - Algorithm selection
KW - Empirical performance estimation
KW - Machine learning
UR - http://www.scopus.com/inward/record.url?scp=84962888054&partnerID=8YFLogxK
U2 - 10.1016/j.artint.2016.04.003
DO - 10.1016/j.artint.2016.04.003
M3 - Article
AN - SCOPUS:84962888054
VL - 237
SP - 41
EP - 58
JO - Artificial intelligence
JF - Artificial intelligence
SN - 0004-3702
ER -