Opening the Black Box: Automated Software Analysis for Algorithm Selection

Damir Pulatov, Marie Anastacio, Lars Kotthoff, Holger Hoos
Proceedings of the First International Conference on Automated Machine Learning, PMLR 188:6/1-18, 2022.

Abstract

Impressive performance improvements have been achieved in many areas of AI by meta-algorithmic techniques, such as automated algorithm selection and configuration. However, existing techniques treat the target algorithms they are applied to as black boxes – nothing is known about their inner workings. This allows meta-algorithmic techniques to be used broadly, but leaves untapped potential performance improvements enabled by information gained from a deeper analysis of the target algorithms. In this paper, we open the black box without sacrificing universal applicability of meta-algorithmic techniques by automatically analyzing algorithms. We show how to use this information to perform algorithm selection, and demonstrate improved performance compared to previous approaches that treat algorithms as black boxes.

Cite this Paper

BibTeX
@InProceedings{pmlr-v188-pulatov22a, title = {Opening the Black Box: Automated Software Analysis for Algorithm Selection}, author = {Pulatov, Damir and Anastacio, Marie and Kotthoff, Lars and Hoos, Holger}, booktitle = {Proceedings of the First International Conference on Automated Machine Learning}, pages = {6/1--18}, year = {2022}, editor = {Guyon, Isabelle and Lindauer, Marius and van der Schaar, Mihaela and Hutter, Frank and Garnett, Roman}, volume = {188}, series = {Proceedings of Machine Learning Research}, month = {25--27 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v188/pulatov22a/pulatov22a.pdf}, url = {https://proceedings.mlr.press/v188/pulatov22a.html}, abstract = {Impressive performance improvements have been achieved in many areas of AI by meta-algorithmic techniques, such as automated algorithm selection and configuration. However, existing techniques treat the target algorithms they are applied to as black boxes – nothing is known about their inner workings. This allows meta-algorithmic techniques to be used broadly, but leaves untapped potential performance improvements enabled by information gained from a deeper analysis of the target algorithms. In this paper, we open the black box without sacrificing universal applicability of meta-algorithmic techniques by automatically analyzing algorithms. We show how to use this information to perform algorithm selection, and demonstrate improved performance compared to previous approaches that treat algorithms as black boxes.} }
Endnote
%0 Conference Paper %T Opening the Black Box: Automated Software Analysis for Algorithm Selection %A Damir Pulatov %A Marie Anastacio %A Lars Kotthoff %A Holger Hoos %B Proceedings of the First International Conference on Automated Machine Learning %C Proceedings of Machine Learning Research %D 2022 %E Isabelle Guyon %E Marius Lindauer %E Mihaela van der Schaar %E Frank Hutter %E Roman Garnett %F pmlr-v188-pulatov22a %I PMLR %P 6/1--18 %U https://proceedings.mlr.press/v188/pulatov22a.html %V 188 %X Impressive performance improvements have been achieved in many areas of AI by meta-algorithmic techniques, such as automated algorithm selection and configuration. However, existing techniques treat the target algorithms they are applied to as black boxes – nothing is known about their inner workings. This allows meta-algorithmic techniques to be used broadly, but leaves untapped potential performance improvements enabled by information gained from a deeper analysis of the target algorithms. In this paper, we open the black box without sacrificing universal applicability of meta-algorithmic techniques by automatically analyzing algorithms. We show how to use this information to perform algorithm selection, and demonstrate improved performance compared to previous approaches that treat algorithms as black boxes.
APA
Pulatov, D., Anastacio, M., Kotthoff, L. & Hoos, H.. (2022). Opening the Black Box: Automated Software Analysis for Algorithm Selection. Proceedings of the First International Conference on Automated Machine Learning, in Proceedings of Machine Learning Research 188:6/1-18 Available from https://proceedings.mlr.press/v188/pulatov22a.html.

Related Material