A Permutation-Equivariant Neural Network Architecture For Auction Design

Designing an incentive compatible auction that maximizes expected revenue is a central problem in Auction Design. Theoretical approaches to the problem have hit some limits in the past decades and analytical solutions are known for only a few simple settings. Computational approaches to the problem through the use of LPs have their own set of limitations. Building on the success of deep learning, a new approach was recently proposed by Duetting et al. (2019) in which the auction is modeled by a feed-forward neural network and the design problem is framed as a learning problem. The neural architectures used in that work are general purpose and do not take advantage of any of the symmetries the problem could present, such as permutation equivariance. In this work, we consider auction design problems that have permutation-equivariant symmetry and construct a neural architecture that is capable of perfectly recovering the permutation-equivariant optimal mechanism, which we show is not possible with the previous architecture. We demonstrate that permutation-equivariant architectures are not only capable of recovering previous results, they also have better generalization properties.

[1]  Pingzhong Tang,et al.  Automated Mechanism Design via Neural Networks , 2018, AAMAS.

[2]  Tim Roughgarden,et al.  Ironing in the Dark , 2015, EC.

[3]  Roger B. Myerson,et al.  Optimal Auction Design , 1981, Math. Oper. Res..

[4]  Tuomas Sandholm,et al.  Automated Design of Revenue-Maximizing Combinatorial Auctions , 2015, Oper. Res..

[5]  Vincent Conitzer,et al.  Computationally Feasible Automated Mechanism Design: General Approach and Case Studies , 2010, AAAI.

[6]  Yannai A. Gonczarowski,et al.  Efficient empirical revenue maximization in single-parameter auction environments , 2016, STOC.

[7]  S. Matthew Weinberg,et al.  Auction learning as a two-player game , 2021, ICLR.

[8]  Jason D. Hartline,et al.  Sample Complexity for Non-Truthful Mechanisms , 2016, EC.

[9]  Yang Cai,et al.  An algorithmic characterization of multi-dimensional mechanisms , 2011, STOC '12.

[10]  Max Welling,et al.  3D Steerable CNNs: Learning Rotationally Equivariant Features in Volumetric Data , 2018, NeurIPS.

[11]  Noam Nisan,et al.  Approximate revenue maximization with multiple items , 2012, EC '12.

[12]  Andrew Chi-Chih Yao,et al.  On revenue maximization for selling multiple independently distributed items , 2013, Proceedings of the National Academy of Sciences.

[13]  Tim Roughgarden,et al.  Simple versus optimal mechanisms , 2009, SECO.

[14]  Christos Tzamos,et al.  The Complexity of Optimal Mechanism Design , 2012, SODA.

[15]  Thore Graepel,et al.  A Neural Architecture for Designing Truthful and Efficient Auctions , 2019, ArXiv.

[16]  Qi Qi,et al.  On the Approximability of Simple Mechanisms for MHR Distributions , 2019, WINE.

[17]  S. Weinberg,et al.  Approximation Schemes for a Unit-Demand Buyer with Independent Items via Symmetries , 2019, 2019 IEEE 60th Annual Symposium on Foundations of Computer Science (FOCS).

[18]  Daniel R. Vincent,et al.  Bayesian and Dominant‐Strategy Implementation in the Independent Private‐Values Model , 2010 .

[19]  Ah Chung Tsoi,et al.  The Graph Neural Network Model , 2009, IEEE Transactions on Neural Networks.

[20]  Christos Tzamos,et al.  Strong Duality for a Multiple-Good Monopolist , 2014, EC.

[21]  Max Welling,et al.  Spherical CNNs , 2018, ICLR.

[22]  Nikhil R. Devanur,et al.  The sample complexity of auctions with side information , 2015, STOC.

[23]  Alexander J. Smola,et al.  Deep Sets , 2017, 1703.06114.

[24]  S. Hart,et al.  Maximal revenue with multiple goods: Nonmonotonicity and other observations , 2015 .

[25]  David C. Parkes,et al.  A General Statistical Framework for Designing Strategy-proof Assignment Mechanisms , 2016, UAI.

[26]  Noam Nisan,et al.  The menu-size complexity of auctions , 2013, EC '13.

[27]  Nikhil R. Devanur,et al.  A duality-based unified approach to Bayesian mechanism design , 2016, SECO.

[28]  Paul Dütting,et al.  Optimal auctions through deep learning , 2017, ICML.

[29]  David C. Parkes,et al.  Deep Learning for Multi-Facility Location Mechanism Design , 2018, IJCAI.

[30]  Yang Cai,et al.  Optimal Multi-dimensional Mechanism Design: Reducing Revenue to Welfare Maximization , 2012, 2012 IEEE 53rd Annual Symposium on Foundations of Computer Science.

[31]  Tim Roughgarden,et al.  On the Pseudo-Dimension of Nearly Optimal Auctions , 2015, NIPS.

[32]  Tim Roughgarden,et al.  Making the Most of Your Samples , 2014, EC.

[33]  Vincent Conitzer,et al.  Complexity of Mechanism Design , 2002, UAI.

[34]  Shuchi Chawla,et al.  Multi-parameter mechanism design and sequential posted pricing , 2010, BQGT.

[35]  S. Matthew Weinberg,et al.  The Sample Complexity of Up-to-ε Multi-Dimensional Revenue Maximization , 2018, 2018 IEEE 59th Annual Symposium on Foundations of Computer Science (FOCS).

[36]  Xi Chen,et al.  On the Complexity of Optimal Lottery Pricing and Randomized Mechanisms , 2015, 2015 IEEE 56th Annual Symposium on Foundations of Computer Science.

[37]  Yoshua Bengio,et al.  Generative Adversarial Nets , 2014, NIPS.

[38]  S. Matthew Weinberg,et al.  Optimal (and benchmark-optimal) competition complexity for additive buyers over independent items , 2019, STOC.

[39]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[40]  Pierre Vandergheynst,et al.  Geometric Deep Learning: Going beyond Euclidean data , 2016, IEEE Signal Process. Mag..

[41]  Tim Roughgarden,et al.  Learning Simple Auctions , 2016, COLT.

[42]  Nima Haghpanah,et al.  The Simple Economics of Approximately Optimal Auctions , 2012, 2013 IEEE 54th Annual Symposium on Foundations of Computer Science.

[43]  Noam Nisan,et al.  Sampling and Representation Complexity of Revenue Maximization , 2014, WINE.

[44]  Richard Cole,et al.  The sample complexity of revenue maximization , 2014, STOC.

[45]  Shuchi Chawla,et al.  Mechanism Design for Subadditive Agents via an Ex Ante Relaxation , 2016, EC.

[46]  Tim Roughgarden,et al.  Revenue maximization with a single sample , 2010, EC '10.

[47]  David C. Parkes,et al.  Deep Learning for Revenue-Optimal Auctions with Budgets , 2018, AAMAS.

[48]  S. Matthew Weinberg,et al.  Symmetries and optimal multi-dimensional mechanism design , 2012, EC '12.

[49]  Saeed Alaei,et al.  Bayesian Combinatorial Auctions: Expanding Single Buyer Mechanisms to Many Buyers , 2011, 2011 IEEE 52nd Annual Symposium on Foundations of Computer Science.

[50]  Vincent Conitzer,et al.  Self-interested automated mechanism design and implications for optimal combinatorial auctions , 2004, EC '04.

[51]  Stewart W. Wilson Knowledge Growth in an Artificial Animal , 1985, ICGA.

[52]  Yang Cai,et al.  Simple mechanisms for subadditive buyers via duality , 2016, STOC.

[53]  Maria-Florina Balcan,et al.  Sample Complexity of Automated Mechanism Design , 2016, NIPS.

[54]  Yang Yuan,et al.  Optimal Auctions vs. Anonymous Pricing , 2015, 2015 IEEE 56th Annual Symposium on Foundations of Computer Science.

[55]  S. Matthew Weinberg,et al.  A Simple and Approximately Optimal Mechanism for an Additive Buyer , 2014, 2014 IEEE 55th Annual Symposium on Foundations of Computer Science.

[56]  Kevin Leyton-Brown,et al.  Deep Models of Interactions Across Sets , 2018, ICML.

[57]  S. Weinberg,et al.  Approximation Schemes for a Buyer with Independent Items via Symmetries , 2019, ArXiv.

[58]  Max Welling,et al.  Group Equivariant Convolutional Networks , 2016, ICML.

[59]  Andrew Chi-Chih Yao,et al.  An n-to-1 Bidder Reduction for Multi-item Auctions and its Applications , 2014, SODA.

[60]  S. Matthew Weinberg,et al.  Simple Mechanisms for a Subadditive Buyer and Applications to Revenue Monotonicity , 2018, ACM Trans. Economics and Comput..

[61]  David Silver,et al.  A Unified Game-Theoretic Approach to Multiagent Reinforcement Learning , 2017, NIPS.

[62]  G. Pavlov,et al.  Optimal Mechanism for Selling Two Goods , 2011 .

[63]  Lawrence D. Jackel,et al.  Backpropagation Applied to Handwritten Zip Code Recognition , 1989, Neural Computation.

[64]  Nima Haghpanah,et al.  Bayesian optimal auctions via multi- to single-agent reduction , 2012, EC '12.

[65]  John Thanassoulis,et al.  Haggling over substitutes , 2004, J. Econ. Theory.

[66]  Max Welling,et al.  Semi-Supervised Classification with Graph Convolutional Networks , 2016, ICLR.

[67]  Pinyan Lu,et al.  Tight approximation ratio of anonymous pricing , 2018, STOC.

[68]  Paul Dütting,et al.  Payment Rules through Discriminant-Based Classifiers , 2012, ACM Trans. Economics and Comput..

[69]  William Vickrey,et al.  Counterspeculation, Auctions, And Competitive Sealed Tenders , 1961 .

[70]  Pinyan Lu,et al.  Tight Revenue Gaps among Simple Mechanisms , 2018, SODA.

[71]  Xi Chen,et al.  On the Complexity of Simple and Optimal Deterministic Mechanisms for an Additive Buyer , 2018, SODA.

[72]  Sébastien Lahaie,et al.  A Kernel-Based Iterative Combinatorial Auction , 2011, AAAI.

[73]  Alejandro M. Manelli,et al.  Bundling as an optimal selling mechanism for a multiple-good monopolist , 2006, J. Econ. Theory.

[74]  Pingzhong Tang,et al.  Optimal mechanisms with simple menus , 2014, EC.

[75]  Barnabás Póczos,et al.  Equivariance Through Parameter-Sharing , 2017, ICML.

[76]  Chenghao Guo,et al.  Settling the sample complexity of single-parameter revenue maximization , 2019, STOC.

[77]  Lantao Yu,et al.  Multi-Agent Adversarial Inverse Reinforcement Learning , 2019, ICML.

[78]  Yaron Lipman,et al.  Invariant and Equivariant Graph Networks , 2018, ICLR.

[79]  Mehryar Mohri,et al.  Learning Theory and Algorithms for revenue optimization in second price auctions with reserve , 2013, ICML.

[80]  International Foundation for Autonomous Agents and MultiAgent Systems ( IFAAMAS ) , 2007 .

[81]  Xi Chen,et al.  The Complexity of Optimal Multidimensional Pricing , 2013, SODA.

[82]  Vasilis Syrgkanis A Sample Complexity Measure with Applications to Learning Optimal Auctions , 2017, NIPS.