Timezone: »

Optimal Auctions through Deep Learning
Paul Duetting · Zhe Feng · Harikrishna Narasimhan · David Parkes · Sai Srivatsa Ravindranath

Tue Jun 11 11:40 AM -- 12:00 PM (PDT) @ Room 102

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 multi-bidder, 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.

Author Information

Paul Duetting (London School of Economics)
Zhe Feng (Harvard University)
Harikrishna Narasimhan (Google Research)
David Parkes (Harvard University)
Sai Srivatsa Ravindranath (Harvard University)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors