An official website of the United States government
The design of optimal auctions is a problem of interest in economics, game theory and computer science. Despite decades of effort, strategyproof, revenue-maximizing auction designs are still not known outside of restricted settings. However, recent methods using deep learning have shown some success in approximating optimal auctions, recovering several known solutions and outperforming strong baselines when optimal auctions are not known. In addition to maximizing revenue, auction mechanisms may also seek to encourage socially desirable constraints such as allocation fairness or diversity. However, these philosophical notions neither have standardization nor do they have widely accepted formal definitions. In this paper, we propose PreferenceNet, an extension of existing neural-network-based auction mechanisms to encode constraints using (potentially human-provided) exemplars of desirable allocations. In addition, we introduce a new metric to evaluate an auction allocations' adherence to such socially desirable constraints and demonstrate that our proposed method is competitive with current state-of-the-art neural-network based auction designs. We validate our approach through human subject research and show that we are able to effectively capture real human preferences.
more »
« less
Learning Revenue-Maximizing Auctions With Differentiable Matching
We propose a new architecture to approximately learn incentive compatible, revenue-maximizing auctions from sampled valuations. Our architecture uses the Sinkhorn algorithm to perform a differentiable bipartite matching which allows the network to learn strategyproof revenue-maximizing mechanisms in settings not learnable by the previous RegretNet architecture. In particular, our architecture is able to learn mechanisms in settings without free disposal where each bidder must be allocated exactly some number of items. In experiments, we show our approach successfully recovers multiple known optimal mechanisms and high-revenue, low-regret mechanisms in larger settings where the optimal mechanism is unknown.
more »
« less
- PAR ID:
- 10409117
- Date Published:
- Journal Name:
- Proceedings of The 25th International Conference on Artificial Intelligence and Statistics
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
null (Ed.)We characterize revenue maximizing mechanisms in a common value environment where the value of the object is equal to the highest of the bidders' independent signals. If the revenue maximizing solution is to sell the object with probability 1, then an optimal mechanism is simply a posted price, namely, the highest price such that every type of every bidder is willing to buy the object. If the object is optimally sold with probability less than 1, then optimal mechanisms skew the allocation toward bidders with lower signals. The resulting allocation induces a “winner's blessing,” whereby the expected value conditional on winning is higher than the unconditional expectation. By contrast, standard auctions that allocate to the bidder with the highest signal (e.g., the first‐price, second‐price, or English auctions) deliver lower revenue because of the winner's curse generated by the allocation. Our qualitative results extend to more general common value environments with a strong winner's curse.more » « less
-
We consider a revenue-maximizing seller with m heterogeneous items and a single buyer whose valuation for the items may exhibit both substitutes and complements. We show that the better of selling the items separately and bundling them together— guarantees a [Formula: see text]-fraction of the optimal revenue, where d is a measure of the degree of complementarity; it extends prior work showing that the same simple mechanism achieves a constant-factor approximation when buyer valuations are subadditive (the most general class of complement-free valuations). Our proof is enabled by a recent duality framework, which we use to obtain a bound on the optimal revenue in the generalized setting. Our technical contributions are domain specific to handle the intricacies of settings with complements. One key modeling contribution is a tractable notion of “degree of complementarity” that admits meaningful results and insights—we demonstrate that previous definitions fall short in this regard.more » « less
-
Most results in revenue-maximizing mechanism design hinge on “getting the price right”—selling goods to bidders at prices low enough to encourage a sale but high enough to garner nontrivial revenue. This approach is difficult to implement when the seller has little or no a priori information about bidder valuations or when the setting is sufficiently complex, such as matching markets with heterogeneous goods. In this paper, we apply a robust approach to designing auctions for revenue. Instead of relying on prior knowledge regarding bidder valuations, we “let the market do the work” and let prices emerge from competition for scarce goods. We analyze the revenue guarantees of one of the simplest imaginable implementations of this idea: first, we enhance competition in the market by increasing demand (or alternatively, by limiting supply), and second, we run a standard second price (Vickrey) auction. In their renowned work from 1996 , Bulow and Klemperer [Bulow J, Klemperer P (1996) Auctions vs. negotiations. Amer. Econom. Rev. 86(1):180–194.] apply this method to markets with single goods. As our main result, we give the first application beyond single-parameter settings, proving that, simultaneously for many valuation distributions, this method achieves expected revenue at least as good as the optimal revenue in the original market. Our robust and simple approach provides a handle on the elusive optimal revenue in multiitem matching markets and shows when the use of welfare-maximizing Vickrey auctions is justified, even if revenue is a priority. By establishing quantitative tradeoffs, our work provides guidelines for a seller in choosing among two different revenue-extracting strategies: sophisticated pricing based on market research or advertising to draw additional bidders.more » « less
-
We provide algorithms that learn simple auctions whose revenue is approximately optimal in multi-item multi-bidder settings, for a wide range of valuations including unit-demand, additive, constrained additive, XOS, and subadditive. We obtain our learning results in two settings. The first is the commonly studied setting where sample access to the bidders' distributions over valuations is given, for both regular distributions and arbitrary distributions with bounded support. Our algorithms require polynomially many samples in the number of items and bidders. The second is a more general max-min learning setting that we introduce, where we are given "approximate distributions," and we seek to compute an auction whose revenue is approximately optimal simultaneously for all "true distributions" that are close to the given ones. These results are more general in that they imply the sample-based results, and are also applicable in settings where we have no sample access to the underlying distributions but have estimated them indirectly via market research or by observation of previously run, potentially non-truthful auctions. Our results hold for valuation distributions satisfying the standard (and necessary) independence-across-items property. They also generalize and improve upon recent works, which have provided algorithms that learn approximately optimal auctions in more restricted settings with additive, subadditive and unit-demand valuations using sample access to distributions. We generalize these results to the complete unit-demand, additive, and XOS setting, to i.i.d. subadditive bidders, and to the max-min setting. Our results are enabled by new uniform convergence bounds for hypotheses classes under product measures. Our bounds result in exponential savings in sample complexity compared to bounds derived by bounding the VC dimension, and are of independent interest.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- The DOI is not currently available.
