Abstract
Algorithm portfolios are known to offer robust performances, efficiently overcoming the weakness of every single algorithm on some particular problem instances. Two complementary approaches to get the best out of an algorithm portfolio are to achieve algorithm selection (AS), and to define a scheduler, sequentially launching a few algorithms on a limited computational budget each. The presented system relies on the joint optimization of a pre-scheduler and a per-instance AS, selecting an algorithm well-suited to the problem instance at hand. ASAP has been thoroughly evaluated against the state-of-the-art during the ICON challenge for algorithm selection, receiving an honorable mention. Its evaluation on several combinatorial optimization benchmarks exposes surprisingly good results of the simple heuristics used; some extensions thereof are presented and discussed in the paper.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
The increase in the overall number of features is handled by an embedded feature selection mechanism, removing all features with negligible importance criterion (<10\(^{-5}\) in the experiments) in a independently learned 10-trees random forest regression model.
- 2.
The codes of all submitted systems and the results are publicly available, http://challenge.icon-fet.eu/challengeas.
- 3.
For CSP-2010 dataset, only two algorithms are available: the pre-scheduler thus consists of a single algorithm, and all ASAP_RF.V2 variants with the same selector hyperparameter are identical.
- 4.
Remind that these instances are not actually passed to the AS in the challenge evaluation setup
References
Bischl, B., Kerschke, P., Kotthoff, L., Lindauer, M., Malitsky, Y., Fréchette, A., et al. (2016). Aslib: a benchmark library for algorithm selection. Artificial Intelligence, 237, 41–58.
Branke, J., Deb, K., Dierolf, H., & Osswald, M. (2004). Finding knees in multi-objective optimization. Parallel problem solving from nature-PPSN VIII (pp. 722–731). Berlin: Springer.
Cauwet, M., Liu, J., Rozière, B., & Teytaud, O. (2016). Algorithm portfolios for noisy optimization. Annals of Mathematics and Artificial Intelligence, 76(1–2), 143–172.
Deb, K. (2003). Multi-objective evolutionary algorithms: introducing bias among pareto-optimal solutions. Advances in evolutionary computing (pp. 263–292). Berlin: Springer.
Gagliolo, M., & Schmidhuber, J. (2011). Algorithm portfolio selection as a bandit problem with unbounded losses. Annals of Mathematics and Artificial Intelligence, 61(2), 49–86.
Gomes, C. P., & Selman, B. (2001). Algorithm portfolios. Artificial Intelligence, 126(1), 43–62.
Hansen, N., Müller, S. D., & Koumoutsakos, P. (2003). Reducing the time complexity of the derandomized evolution strategy with covariance matrix adaptation. Evolutionary Computation, 11(1), 1–18.
Huberman, B. A., Lukose, R. M., & Hogg, T. (1997). An economics approach to hard computational problems. Science, 275(5296), 51–54.
Kadioglu, S., Malitsky, Y., Sabharwal, A., Samulowitz, H., & Sellmann, M. (2011). Algorithm selection and scheduling. In Proceedings of the 17th CP (pp. 454–469). LNCS. Berlin: Springer.
Kotthoff, L. (2015). ICON challenge on algorithm selection. CoRR arXiv:abs/1511.04326.
Kotthoff, L. (2016). Algorithm selection for combinatorial search problems: a survey (pp. 149–190). Berlin: Springer International Publishing.
Leyton-Brown, K., Nudelman, E., Andrew, G., McFadden, J., & Shoham, Y. (2003). A portfolio approach to algorithm selection. In Proceedings of the IJCAI (pp. 1542–1543).
Lindauer, M., Bergdoll, R. D., & Hutter, F. (2016). An empirical study of per-instance algorithm scheduling. In Proceedings of the LION10 (pp. 253–259). Berlin: Springer.
Malitsky, Y., Sabharwal, A., & Samulowitz, H. (2013). Algorithm portfolios based on cost-sensitive hierarchical clustering. In Proceedings of the 23rd IJCAI (pp. 608–614). California: AAAI Press.
Mısır, M., & Sebag, M. (2017). Alors: an algorithm recommender system. Artificial Intelligence, 244, 291–314. Published on-line Dec. 2016.
O’Mahony, E., Hebrard, E., Holland, A., Nugent, C., & O’Sullivan, B. (2008). Using case-based reasoning in an algorithm portfolio for constraint solving. In Proceedings of the ICAICS (pp. 210–216).
Oentaryo, R. J., Handoko, S. D., & Lau, H. C. (2015). Algorithm selection via ranking. In Proceedings of the 29th AAAI Conference on Artificial Intelligence (AAAI).
Pedregosa, F., et al. (2011). Scikit-learn: machine learning in python. Journal of Machine Learning Research, 12, 2825–2830.
Rice, J. R. (1976). The algorithm selection problem. Advances in Computers, 15, 65–118.
Stern, D., Herbrich, R., Graepel, T., Samulowitz, H., Pulina, L., & Tacchella, A. (2010). Collaborative expert portfolio management. In Proceedings of the 24th AAAI (pp. 179–184).
Wolpert, D., & Macready, W. (1997). No free lunch theorems for optimization. IEEE Transactions on Evolutionary Computation, 1(1), 67–82.
Xu, L., Hutter, F., Hoos, H., & Leyton-Brown, K. (2012). Features for SAT. University of British Columbia.
Xu, L., Hutter, F., & Hoos, H. H. (2008). Satzilla: portfolio-based algorithm selection for sat. Journal of Artificial Intelligence Research, pp. 565–606.
Xu, L., Hutter, F., Shen, J., Hoos, H. H., & Leyton-Brown, K. (2012). Satzilla2012: improved algorithm selection based on cost-sensitive classification models. In Proceedings of the SAT Challenge (pp. 57–58).
Yun, X., & Epstein, S. L. (2012). Learning algorithm portfolios for parallel execution. In Y. Hamadi & M. Schoenauer (Eds.), In Proceedings of the LION 6 (Vol. 7219, pp. 323–338). LNCS. Berlin: Springer.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2019 Springer International Publishing AG, part of Springer Nature
About this chapter
Cite this chapter
Gonard, F., Schoenauer, M., Sebag, M. (2019). Algorithm Selector and Prescheduler in the ICON Challenge. In: Talbi, EG., Nakib, A. (eds) Bioinspired Heuristics for Optimization. Studies in Computational Intelligence, vol 774. Springer, Cham. https://doi.org/10.1007/978-3-319-95104-1_13
Download citation
DOI: https://doi.org/10.1007/978-3-319-95104-1_13
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-95103-4
Online ISBN: 978-3-319-95104-1
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)