Abstract
The area of combinatorial optimization is characterized by the search for optimal combinations of discrete variables that satisfy some set of constraints. Famous problems in this space include maximum satisfiability and maximum independent set. Due to their discrete dynamics, these problems are not differentiable in their natural formulations. In this paper, we explore the counter-intuitive direction of differentiable discrete optimization by leveraging the recently discovered dataless neural networks, which have been used to yield a single differentiable function that is equivalent to the maximum independent set problem. In particular, we leverage the dataless neural networks framework to derive differentiable forms for a variety of NP-hard discrete problems and prove the correctness of our derivations. The proposed differentiable forms open up the avenue for continuous differentiable optimization to be brought to bear on classical discrete optimization problems.
K. Subramani and S. K. Jena—This research was supported in part by the Defense Advanced Research Projects Agency through grant HR001123S0001-FP-004.
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
Similar content being viewed by others
References
Akiba, T., Iwata, Y.: Branch-and-reduce exponential/FPT algorithms in practice: a case study of vertex cover. Theoret. Comput. Sci. 609, 211–225 (2016)
Alkhouri, I.R., Atia, G.K., Velasquez, A.: A differentiable approach to the maximum independent set problem using dataless neural networks. Neural Netw. 155, 168–176 (2022)
Bengio, Y., Lodi, A., Prouvost, A.: Machine learning for combinatorial optimization: a methodological tour d’horizon. Eur. J. Oper. Res. 290(2), 405–421 (2021)
Boppana, R., Halldórsson, M.M.: Approximating maximum independent sets by excluding subgraphs. BIT Numer. Math. 32(2), 180–196 (1992)
Cygan, M., et al.: Parameterized Algorithms. Springer, Cham (2015). https://doi.org/10.1007/978-3-319-21275-3
Drori, I., et al.: Learning to solve combinatorial optimization problems on real-world graphs in linear time. In: 2020 19th IEEE International Conference on Machine Learning and Applications (ICMLA), pp. 19–24 (2020)
Festa, P.: A brief introduction to exact, approximation, and heuristic algorithms for solving hard combinatorial optimization problems. In: 2014 16th International Conference on Transparent Optical Networks (ICTON), pp. 1–20 (2014)
Flum, J., Grohe, M.: Parameterized Complexity Theory. TTCSAES, Springer, Heidelberg (2006). https://doi.org/10.1007/3-540-29953-X
Fomin, F.V., Kratsch, D.: Exact Exponential Algorithms. TTCSAES, Springer, Heidelberg (2010). https://doi.org/10.1007/978-3-642-16533-7
Gaspers, S.: Exponential Time Algorithms - Structures, Measures, and Bounds. VDM (2010)
Lamm, S., Sanders, P., Schulz, C., Strash, D., Werneck, R.F.: Finding near-optimal independent sets at scale. In: 2016 Proceedings of the Eighteenth Workshop on Algorithm Engineering and Experiments (ALENEX), pp. 138–150 (2016)
Li, Z., Chen, Q., Koltun, V.: Combinatorial optimization with graph convolutional networks and guided tree search. In: Advances in Neural Information Processing Systems, vol. 31 (2018)
Mazyavkina, N., Sviridov, S., Ivanov, S., Burnaev, E.: Reinforcement learning for combinatorial optimization: a survey. Comput. Oper. Res. 134, 105400 (2021)
Niedermeier, R.: Invitation to Fixed-Parameter Algorithms. Oxford University Press (2006)
Segundo, P.S., RodrÃguez-Losada, D., Jiménez, A.: An exact bit-parallel algorithm for the maximum clique problem. Comput. Oper. Res. 38(2), 571–581 (2011)
Schuetz, M.J.A., Brubaker, J.K., Katzgraber, H.G.: Combinatorial optimization with physics-inspired graph neural networks. Nat. Mach. Intell. 4(4), 367–377 (2022)
Wilder, B., Dilkina, B., Tambe, M.: Melding the data-decisions pipeline: Decision-focused learning for combinatorial optimization. In: Proceedings of the AAAI Conference on Artificial Intelligence, vol. 33, pp. 1658–1665 (2019)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2024 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Jena, S.K., Subramani, K., Velasquez, A. (2024). Differentiable Discrete Optimization Using Dataless Neural Networks. In: Wu, W., Guo, J. (eds) Combinatorial Optimization and Applications. COCOA 2023. Lecture Notes in Computer Science, vol 14462. Springer, Cham. https://doi.org/10.1007/978-3-031-49614-1_1
Download citation
DOI: https://doi.org/10.1007/978-3-031-49614-1_1
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-49613-4
Online ISBN: 978-3-031-49614-1
eBook Packages: Computer ScienceComputer Science (R0)