research-article

Free access

Customizing ML predictions for online algorithms

AUTHORs:

Debmalya PanigrahiAuthors Info & Claims

ICML'20: Proceedings of the 37th International Conference on Machine Learning

Article No.: 29, Pages 303 - 313

Published: 13 July 2020 Publication History

PDF eReader Publisher Site

Abstract

A popular line of recent research incorporates ML advice in the design of online algorithms to improve their performance in typical instances. These papers treat the ML algorithm as a black-box, and redesign online algorithms to take advantage of ML predictions. In this paper, we ask the complementary question: can we redesign ML algorithms to provide better predictions for on-line algorithms? We explore this question in the context of the classic rent-or-buy problem, and show that incorporating optimization benchmarks in ML loss functions leads to signifcantly better performance, while maintaining a worst-case adversarial result when the advice is completely wrong. We support this fnding both through theoretical bounds and numerical simulations.

References

[1]

Ailon, N., Chazelle, B., Clarkson, K. L., Liu, D., Mulzer, W., and Seshadhri, C. Self-improving algorithms. SIAM Journal on Computing, 40(2):350-375, 2011.

Digital Library

[2]

Awasthi, P., Balcan, M. F., and Long, P. M. The power of localization for effciently learning linear separators with noise. In Proceedings of the forty-sixth annual ACM symposium on Theory of computing, pp. 449-458. ACM, 2014.

Digital Library

[3]

Balkanski, E., Rubinstein, A., and Singer, Y. The power of optimization from samples. In Lee, D. D., Sugiyama, M., von Luxburg, U., Guyon, I., and Garnett, R. (eds.), Advances in Neural Information Processing Systems 29: Annual Conference on Neural Information Processing Systems 2016, December 5-10, 2016, Barcelona, Spain, pp. 4017-4025, 2016.

[4]

Balkanski, E., Rubinstein, A., and Singer, Y. The limitations of optimization from samples. In Hatami, H., McKenzie, P., and King, V. (eds.), Proceedings of the 49th Annual ACM SIGACT Symposium on Theory of Computing, STOC 2017, Montreal, QC, Canada, June 19-23, 2017, pp. 1016-1027. ACM, 2017.

Digital Library

[5]

Blum, A., Frieze, A., Kannan, R., and Vempala, S. A polynomial-time algorithm for learning noisy linear threshold functions. Algorithmica, 22(1-2):35-52, 1998.

[6]

Bylander, T. Learning linear threshold functions in the presence of classifcation noise. In Proceedings of the seventh annual conference on Computational learning theory, pp. 340-347, 1994.

Digital Library

[7]

Cole, R. and Roughgarden, T. The sample complexity of revenue maximization. In Shmoys, D. B. (ed.), Symposium on Theory of Computing, STOC 2014, New York, NY, USA, May 31 -June 03, 2014, pp. 243-252. ACM, 2014.

[8]

Elkan, C. The foundations of cost-sensitive learning. In International joint conference on artifcial intelligence, volume 17, pp. 973-978. Lawrence Erlbaum Associates Ltd, 2001.

[9]

Gollapudi, S. and Panigrahi, D. Online algorithms for rentor-buy with expert advice. In Proceedings of the 36th International Conference on Machine Learning, ICML 2019, 9-15 June 2019, Long Beach, California, USA, pp. 2319-2327, 2019.

[10]

Gupta, R. and Roughgarden, T. A pac approach to application-specifc algorithm selection. SIAM Journal on Computing, 46(3):992-1017, 2017.

[11]

Hsu, C.-Y., Indyk, P., Katabi, D., and Vakilian, A. Learning-based frequency estimation algorithms. In International Conference on Learning Representations, 2019. URL https://openreview.net/forum? id=r1lohoCqY7.

[12]

Huang, C., Zhai, S., Talbott, W., Bautista, M. A., Sun, S.-Y., Guestrin, C., and Susskind, J. Addressing the loss-metric mismatch with adaptive loss alignment. arXiv preprint arXiv:1905.05895, 2019.

[13]

Jiang, Z., Panigrahi, D., and Sun, K. Online algorithms for weighted caching with predictions. In 47th International Colloquium on Automata, Languages, and Programming, ICALP 2020, 2020.

[14]

Kamalaruban, P. and Williamson, R. C. Minimax lower bounds for cost sensitive classifcation. arXiv preprint arXiv:1805.07723, 2018.

[15]

Karlin, A. R., Manasse, M. S., Rudolph, L., and Sleator, D. D. Competitive snoopy caching. Algorithmica, 3: 77-119, 1988.

Digital Library

[16]

Karlin, A. R., Manasse, M. S., McGeoch, L. A., and Owicki, S. Competitive randomized algorithms for nonuniform problems. Algorithmica, 11(6):542-571, 1994.

Digital Library

[17]

Karlin, A. R., Kenyon, C., and Randall, D. Dynamic TCP acknowledgment and other stories about e/(e-1). Algorithmica, 36(3):209-224, 2003.

Digital Library

[18]

Kearns, M. J. and Vazirani, U. V. An introduction to computational learning theory. MIT press, 1994.

[19]

Khanafer, A., Kodialam, M., and Puttaswamy, K. P. N. The constrained ski-rental problem and its application to on-line cloud cost optimization. In Proceedings of the INFOCOM, pp. 1492-1500, 2013.

[20]

Kodialam, R. Competitive algorithms for an online rent or buy problem with variable demand. SIAM Undergraduate Research Online, 7:233-245, 2014.

[21]

Lattanzi, S., Lavastida, T., Moseley, B., and Vassilvitskii, S. Online scheduling via learned weights. In Chawla, S. (ed.), Proceedings of the 2020 ACM-SIAM Symposium on Discrete Algorithms, SODA 2020, Salt Lake City, UT, USA, January 5-8, 2020, pp. 1859-1877. SIAM, 2020.

[22]

Ling, C. X. and Sheng, V. S. Cost-sensitive learning and the class imbalance problem, 2008.

[23]

Lotker, Z., Patt-Shamir, B., and Rawitz, D. Rent, lease or buy: Randomized algorithms for multislope ski rental. In Proceedings of the 25th Annual Symposium on the Theoretical Aspects of Computer Science (STACS), pp. 503-514, 2008.

[24]

Lykouris, T. and Vassilvitskii, S. Competitive caching with machine learned advice. arXiv preprint arXiv:1802.05399, 2018.

[25]

Medina, A. M. and Vassilvitskii, S. Revenue optimization with approximate bid predictions. In Advances in Neural Information Processing Systems 30: Annual Conference on Neural Information Processing Systems 2017, 4-9 December 2017, Long Beach, CA, USA, pp. 1858-1866, 2017.

[26]

Meyerson, A. The parking permit problem. In Proc. of 46th Annual IEEE Symposium on Foundations of Computer Science (FOCS), pp. 274-284, 2005.

Digital Library

[27]

Mitzenmacher, M. A model for learned bloom flters and optimizing by sandwiching. In Advances in Neural Information Processing Systems, pp. 464-473, 2018.

[28]

Mitzenmacher, M. Scheduling with predictions and the price of misprediction. In Vidick, T. (ed.), 11th Innova-tions in Theoretical Computer Science Conference, ITCS 2020, January 12-14, 2020, Seattle, Washington, USA, volume 151 of LIPIcs, pp. 14:1-14:18. Schloss Dagstuhl -Leibniz-Zentrum fur Informatik, 2020.

[29]

Morgenstern, J. and Roughgarden, T. Learning simple auctions. In Feldman, V., Rakhlin, A., and Shamir, O. (eds.), Proceedings of the 29th Conference on Learning Theory, COLT 2016, New York, USA, June 23-26, 2016, volume 49 of JMLR Workshop and Conference Proceedings, pp. 1298-1318. JMLR.org, 2016.

[30]

Motwani, R. and Raghavan, P. Randomized Algorithms. Cambridge University Press, 1997.

Digital Library

[31]

Natarajan, N., Dhillon, I. S., Ravikumar, P. K., and Tewari, A. Learning with noisy labels. In Advances in neural information processing systems, pp. 1196-1204, 2013.

Digital Library

[32]

Purohit, M., Svitkina, Z., and Kumar, R. Improving online algorithms via ml predictions. In Advances in Neural Information Processing Systems, pp. 9661-9670, 2018.

[33]

Rasmussen, C. E. Gaussian processes in machine learning. In Summer School on Machine Learning, pp. 63-71. Springer, 2003.

[34]

Rohatgi, D. Near-optimal bounds for online caching with machine learned advice. In Chawla, S. (ed.), Proceedings of the 2020 ACM-SIAM Symposium on Discrete Algorithms, SODA 2020, Salt Lake City, UT, USA, January 5-8, 2020, pp. 1834-1845. SIAM, 2020.

[35]

Vapnik, V. and Vapnik, V. Statistical learning theory wiley. New York, pp. 156-160, 1998.

Cited By

Khodak MAmin KDick TVassilvitskii SKrause ABrunskill ECho KEngelhardt BSabato SScarlett J(2023)Learning-augmented private algorithms for multiple quantile releaseProceedings of the 40th International Conference on Machine Learning10.5555/3618408.3619078(16344-16376)Online publication date: 23-Jul-2023
https://dl.acm.org/doi/10.5555/3618408.3619078
Khodak MBalcan MTalwalkar AVassilvitskii SKoyejo SMohamed SAgarwal ABelgrave DCho KOh A(2022)Learning predictions for algorithms with predictionsProceedings of the 36th International Conference on Neural Information Processing Systems10.5555/3600270.3600526(3542-3555)Online publication date: 28-Nov-2022
https://dl.acm.org/doi/10.5555/3600270.3600526

Index Terms

Customizing ML predictions for online algorithms
1. Mathematics of computing
2. Theory of computation

Index terms have been assigned to the content through auto-classification.

Recommendations

Improving online algorithms via ML predictions
NIPS'18: Proceedings of the 32nd International Conference on Neural Information Processing Systems

In this work we study the problem of using machine-learned predictions to improve the performance of online algorithms. We consider two classical problems, ski rental and non-clairvoyant job scheduling, and obtain new online algorithms that use ...
Online Algorithms with Predictions
Multigroup ML decodable collocated and distributed space-time block codes

In this paper, collocated and distributed space-time block codes (DSTBCs) which admit multigroup maximum-likelihood (ML) decoding are studied. First, the collocated case is considered and the problem of constructing space-time block codes (STBCs) which ...

Comments

Information & Contributors

Information

Published In

cover image Guide Proceedings

ICML'20: Proceedings of the 37th International Conference on Machine Learning

July 2020

11702 pages

Editors:
Hal Daumé,
Aarti Singh

Copyright © 2020.

Publisher

JMLR.org

Publication History

Published: 13 July 2020

Qualifiers

Research-article
Research
Refereed limited

Contributors

Other Metrics

View Article Metrics

Bibliometrics & Citations

Bibliometrics

Article Metrics

2
Total Citations
View Citations
36
Total Downloads

Downloads (Last 12 months)28
Downloads (Last 6 weeks)7

Reflects downloads up to 12 Aug 2024

Other Metrics

View Author Metrics

Citations

Cited By

Khodak MAmin KDick TVassilvitskii SKrause ABrunskill ECho KEngelhardt BSabato SScarlett J(2023)Learning-augmented private algorithms for multiple quantile releaseProceedings of the 40th International Conference on Machine Learning10.5555/3618408.3619078(16344-16376)Online publication date: 23-Jul-2023
https://dl.acm.org/doi/10.5555/3618408.3619078
Khodak MBalcan MTalwalkar AVassilvitskii SKoyejo SMohamed SAgarwal ABelgrave DCho KOh A(2022)Learning predictions for algorithms with predictionsProceedings of the 36th International Conference on Neural Information Processing Systems10.5555/3600270.3600526(3542-3555)Online publication date: 28-Nov-2022
https://dl.acm.org/doi/10.5555/3600270.3600526

View Options

View options

PDF

View or Download as a PDF file.

eReader

View online with eReader.

Get Access

Login options

Check if you have access through your login credentials or your institution to get full access on this article.

Full Access

Get this Publication

Media

Figures

Other

Tables

View Table of Contents