Optimal Auctions through Deep Learning

Paul Duetting
Zhe Feng
David C. Parkes
Sai Srivatsa Ravindranath
Proceedings of the 36th International Conference on Machine Learning, PMLR, Long Beach, California, USA(2019), pp. 1706-1715


Designing an incentive compatible auction that maximizes expected revenue is an intricate task. The single-item case was resolved in a seminal piece of work by Myerson in 1981. Even after 30-40 years of intense research the problem remains unsolved for seemingly simple multibidder, multi-item settings. In this work, we initiate the exploration of the use of tools from deep learning for the automated design of optimal auctions. We model an auction as a multi-layer neural network, frame optimal auction design as a constrained learning problem, and show how it can be solved using standard pipelines. We prove generalization bounds and present extensive experiments, recovering essentially all known analytical solutions for multi-item settings, and obtaining novel mechanisms for settings in which the optimal mechanism is unknown.