Loading [MathJax]/extensions/tex2jax.js

Efficient Parameter Importance Analysis via Ablation with Surrogates

Research output: Chapter in book/report/conference proceedingConference contributionResearchpeer review

Authors

  • André Biedenkapp
  • Marius Lindauer
  • Katharina Eggensperger
  • Frank Hutter

External Research Organisations

  • University of Freiburg
  • University of British Columbia

Details

Original languageEnglish
Title of host publicationProceedings of the AAAI Conference on Artificial Intelligence
Publication statusPublished - 2 Nov 2017
Externally publishedYes

Abstract

To achieve peak performance, it is often necessary to adjust the parameters of a given algorithm to the class of problem instances to be solved; this is known to be the case for popular solvers for a broad range of AI problems, including AI planning, propositional satisfiability (SAT) and answer set programming (ASP). To avoid tedious and often highly sub-optimal manual tuning of such parameters by means of ad-hoc methods, general-purpose algorithm configuration procedures can be used to automatically find performance-optimizing parameter settings. While impressive performance gains are often achieved in this manner, additional, potentially costly parameter importance analysis is required to gain insights into what parameter changes are most responsible for those improvements. Here, we show how the running time cost of ablation analysis, a wellknown general-purpose approach for assessing parameter importance, can be reduced substantially by using regression models of algorithm performance constructed from data collected during the configuration process. In our experiments, we demonstrate speed-up factors between 33 and 14 727 for ablation analysis on various configuration scenarios from AI planning, SAT, ASP and mixed integer programming (MIP).

ASJC Scopus subject areas

Cite this

Efficient Parameter Importance Analysis via Ablation with Surrogates. / Biedenkapp, André; Lindauer, Marius; Eggensperger, Katharina et al.
Proceedings of the AAAI Conference on Artificial Intelligence. 2017.

Research output: Chapter in book/report/conference proceedingConference contributionResearchpeer review

Biedenkapp, A, Lindauer, M, Eggensperger, K, Hutter, F, Fawcett, C & Hoos, HH 2017, Efficient Parameter Importance Analysis via Ablation with Surrogates. in Proceedings of the AAAI Conference on Artificial Intelligence. https://doi.org/10.1609/aaai.v31i1.10657
Biedenkapp, A., Lindauer, M., Eggensperger, K., Hutter, F., Fawcett, C., & Hoos, H. H. (2017). Efficient Parameter Importance Analysis via Ablation with Surrogates. In Proceedings of the AAAI Conference on Artificial Intelligence https://doi.org/10.1609/aaai.v31i1.10657
Biedenkapp A, Lindauer M, Eggensperger K, Hutter F, Fawcett C, Hoos HH. Efficient Parameter Importance Analysis via Ablation with Surrogates. In Proceedings of the AAAI Conference on Artificial Intelligence. 2017 Epub 2017 Feb 12. doi: 10.1609/aaai.v31i1.10657
Biedenkapp, André ; Lindauer, Marius ; Eggensperger, Katharina et al. / Efficient Parameter Importance Analysis via Ablation with Surrogates. Proceedings of the AAAI Conference on Artificial Intelligence. 2017.
Download
@inproceedings{7684341d56824ccd99ae9cf107e22dca,
title = "Efficient Parameter Importance Analysis via Ablation with Surrogates",
abstract = "To achieve peak performance, it is often necessary to adjust the parameters of a given algorithm to the class of problem instances to be solved; this is known to be the case for popular solvers for a broad range of AI problems, including AI planning, propositional satisfiability (SAT) and answer set programming (ASP). To avoid tedious and often highly sub-optimal manual tuning of such parameters by means of ad-hoc methods, general-purpose algorithm configuration procedures can be used to automatically find performance-optimizing parameter settings. While impressive performance gains are often achieved in this manner, additional, potentially costly parameter importance analysis is required to gain insights into what parameter changes are most responsible for those improvements. Here, we show how the running time cost of ablation analysis, a wellknown general-purpose approach for assessing parameter importance, can be reduced substantially by using regression models of algorithm performance constructed from data collected during the configuration process. In our experiments, we demonstrate speed-up factors between 33 and 14 727 for ablation analysis on various configuration scenarios from AI planning, SAT, ASP and mixed integer programming (MIP).",
author = "Andr{\'e} Biedenkapp and Marius Lindauer and Katharina Eggensperger and Frank Hutter and Chris Fawcett and Hoos, {Holger H.}",
note = "Funding information: Acknowledgments K. Eggensperger, M. Lindauer and F. Hutter acknowledge funding by the DFG (German Research Foundation) under Emmy Noether grant HU 1900/2-1; K. Eggensperger also acknowledges funding by the State Graduate Funding Program of Baden-W{\"u}rttemberg. H. Hoos acknowledges funding through an NSERC Discovery Grant.",
year = "2017",
month = nov,
day = "2",
doi = "10.1609/aaai.v31i1.10657",
language = "English",
booktitle = "Proceedings of the AAAI Conference on Artificial Intelligence",

}

Download

TY - GEN

T1 - Efficient Parameter Importance Analysis via Ablation with Surrogates

AU - Biedenkapp, André

AU - Lindauer, Marius

AU - Eggensperger, Katharina

AU - Hutter, Frank

AU - Fawcett, Chris

AU - Hoos, Holger H.

N1 - Funding information: Acknowledgments K. Eggensperger, M. Lindauer and F. Hutter acknowledge funding by the DFG (German Research Foundation) under Emmy Noether grant HU 1900/2-1; K. Eggensperger also acknowledges funding by the State Graduate Funding Program of Baden-Württemberg. H. Hoos acknowledges funding through an NSERC Discovery Grant.

PY - 2017/11/2

Y1 - 2017/11/2

N2 - To achieve peak performance, it is often necessary to adjust the parameters of a given algorithm to the class of problem instances to be solved; this is known to be the case for popular solvers for a broad range of AI problems, including AI planning, propositional satisfiability (SAT) and answer set programming (ASP). To avoid tedious and often highly sub-optimal manual tuning of such parameters by means of ad-hoc methods, general-purpose algorithm configuration procedures can be used to automatically find performance-optimizing parameter settings. While impressive performance gains are often achieved in this manner, additional, potentially costly parameter importance analysis is required to gain insights into what parameter changes are most responsible for those improvements. Here, we show how the running time cost of ablation analysis, a wellknown general-purpose approach for assessing parameter importance, can be reduced substantially by using regression models of algorithm performance constructed from data collected during the configuration process. In our experiments, we demonstrate speed-up factors between 33 and 14 727 for ablation analysis on various configuration scenarios from AI planning, SAT, ASP and mixed integer programming (MIP).

AB - To achieve peak performance, it is often necessary to adjust the parameters of a given algorithm to the class of problem instances to be solved; this is known to be the case for popular solvers for a broad range of AI problems, including AI planning, propositional satisfiability (SAT) and answer set programming (ASP). To avoid tedious and often highly sub-optimal manual tuning of such parameters by means of ad-hoc methods, general-purpose algorithm configuration procedures can be used to automatically find performance-optimizing parameter settings. While impressive performance gains are often achieved in this manner, additional, potentially costly parameter importance analysis is required to gain insights into what parameter changes are most responsible for those improvements. Here, we show how the running time cost of ablation analysis, a wellknown general-purpose approach for assessing parameter importance, can be reduced substantially by using regression models of algorithm performance constructed from data collected during the configuration process. In our experiments, we demonstrate speed-up factors between 33 and 14 727 for ablation analysis on various configuration scenarios from AI planning, SAT, ASP and mixed integer programming (MIP).

UR - http://www.scopus.com/inward/record.url?scp=85026381970&partnerID=8YFLogxK

U2 - 10.1609/aaai.v31i1.10657

DO - 10.1609/aaai.v31i1.10657

M3 - Conference contribution

BT - Proceedings of the AAAI Conference on Artificial Intelligence

ER -

By the same author(s)