arXiv subjects

Yang Cai

Publications and source records attributed to Yang Cai.

At least 91 records · Page 5Linked to original sources

arXiv2020

Simple Mechanisms for Profit Maximization in Multi-item Auctions

We study a classical Bayesian mechanism design problem where a seller is selling multiple items to multiple buyers. We consider the case where the seller has costs to produce the items, and these costs are private information to the seller. How can the seller design a mechanism to maximize her profit? Two well-studied problems, revenue maximization in multi-item auctions and signaling in ad auctions, are special cases of our problem. We show that there exists a simple mechanism whose profit is at least $\frac{1}{44}$ of the optimal profit for multiple buyers with matroid-rank valuation functions. When there is a single buyer, the approximation factor is $11$ for general constraint-additive valuations and $6$ for additive valuations. Our result holds even when the seller's costs are correlated across items. We introduce a new class of mechanisms called permit-selling mechanisms. For single buyer case these mechanisms are quite simple: there are two stages. For each item $j$, we create a separate permit that allows the buyer to purchase the item in the second stage. In the first stage, we sell the permits without revealing any information about the costs. In the second stage, the seller reveals all the costs, and the buyer can buy item $j$ by paying the item price if the buyer has purchased the permit for item $j$ in the first stage. We show that either selling the permits separately or as a grand bundle suffices to achieve the constant factor approximation to the optimal profit (6 for additive, and 11 for constrained additive). For multiple buyers, we sell the permits sequentially and obtain the constant factor approximation. Our proof is enabled by constructing a benchmark for the optimal profit by combining a novel dual solution with the existing ex-ante relaxation technique.

cs.GT↗

arXiv2020

Third-Party Data Providers Ruin Simple Mechanisms

Motivated by the growing prominence of third-party data providers in online marketplaces, this paper studies the impact of the presence of third-party data providers on mechanism design. When no data provider is present, it has been shown that simple mechanisms are "good enough" -- they can achieve a constant fraction of the revenue of optimal mechanisms. The results in this paper demonstrate that this is no longer true in the presence of a third-party data provider who can provide the bidder with a signal that is correlated with the item type. Specifically, even with a single seller, a single bidder, and a single item of uncertain type for sale, the strategies of pricing each item-type separately (the analog of item pricing for multi-item auctions) and bundling all item-types under a single price (the analog of grand bundling) can both simultaneously be a logarithmic factor worse than the optimal revenue. Further, in the presence of a data provider, item-type partitioning mechanisms---a more general class of mechanisms which divide item-types into disjoint groups and offer prices for each group---still cannot achieve within a $\log \log$ factor of the optimal revenue. Thus, our results highlight that the presence of a data-provider forces the use of more complicated mechanisms in order to achieve a constant fraction of the optimal revenue.

cs.GT↗

arXiv2019

Simple and Approximately Optimal Pricing for Proportional Complementarities

We study a new model of complementary valuations, which we call "proportional complementarities." In contrast to common models, such as hypergraphic valuations, in our model, we do not assume that the extra value derived from owning a set of items is independent of the buyer's base valuations for the items. Instead, we model the complementarities as proportional to the buyer's base valuations, and these proportionalities are known market parameters. Our goal is to design a simple pricing scheme that, for a single buyer with proportional complementarities, yields approximately optimal revenue. We define a new class of mechanisms where some number of items are given away for free, and the remaining items are sold separately at inflated prices. We find that the better of such a mechanism and selling the grand bundle earns a 12-approximation to the optimal revenue for pairwise proportional complementarities. This confirms the intuition that items should not be sold completely separately in the presence of complementarities. In the more general case, a buyer has a maximum of proportional positive hypergraphic valuations, where a hyperedge in a given hypergraph describes the boost to the buyer's value for item $i$ given by owning any set of items $T$ in addition. The maximum-out-degree of such a hypergraph is $d$, and $k$ is the positive rank of the hypergraph. For valuations given by these parameters, our simple pricing scheme is an $O(\min\{d,k\})$-approximation.

cs.GT↗

arXiv2019

Simple Mechanisms for Subadditive Buyers via Duality

We provide simple and approximately revenue-optimal mechanisms in the multi-item multi-bidder settings. We unify and improve all previous results, as well as generalize the results to broader cases. In particular, we prove that the better of the following two simple, deterministic and Dominant Strategy Incentive Compatible mechanisms, a sequential posted price mechanism or an anonymous sequential posted price mechanism with entry fee, achieves a constant fraction of the optimal revenue among all randomized, Bayesian Incentive Compatible mechanisms, when buyers' valuations are XOS over independent items. If the buyers' valuations are subadditive over independent items, the approximation factor degrades to $O(\log m)$, where $m$ is the number of items. We obtain our results by first extending the Cai-Devanur-Weinberg duality framework to derive an effective benchmark of the optimal revenue for subadditive bidders, and then analyzing this upper bound with new techniques.

cs.GT↗

arXiv2018

A Duality-Based Unified Approach to Bayesian Mechanism Design

We provide a unified view of many recent developments in Bayesian mechanism design, including the black-box reductions of Cai et al. [CDW13b], simple auctions for additive buyers [HN12], and posted-price mechanisms for unit-demand bidders [CHK07]. Additionally, we show that viewing these three previously disjoint lines of work through the same lens leads to new developments as well. First, we provide a duality framework for Bayesian mechanism design, which naturally accommodates multiple agents and arbitrary objectives/feasibility constraints. Using this, we prove that either a posted-price mechanism or the Vickrey-Clarke-Groves auction with per-bidder entry fees achieves a constant-factor of the optimal revenue achievable by a Bayesian Incentive Compatible mechanism whenever buyers are unit-demand or additive, unifying previous breakthroughs of Chawla et al. [CHMS10] and Yao [Yao15], and improving both approximation ratios (from 30 to 24 and 69 to 8, respectively). Finally, we show that this view also leads to improved structural characterizations in the Cai et al. framework.

cs.GT↗

arXiv2018

Learning Safe Policies with Expert Guidance

We propose a framework for ensuring safe behavior of a reinforcement learning agent when the reward function may be difficult to specify. In order to do this, we rely on the existence of demonstrations from expert policies, and we provide a theoretical framework for the agent to optimize in the space of rewards consistent with its existing knowledge. We propose two methods to solve the resulting optimization: an exact ellipsoid-based method and a method in the spirit of the "follow-the-perturbed-leader" algorithm. Our experiments demonstrate the behavior of our algorithm in both discrete and continuous problems. The trained agent safely avoids states with potential negative effects while imitating the behavior of the expert in the other states.

cs.LG↗

arXiv2018

The Best of Both Worlds: Asymptotically Efficient Mechanisms with a Guarantee on the Expected Gains-From-Trade

The seminal impossibility result of Myerson and Satterthwaite (1983) states that for bilateral trade, there is no mechanism that is individually rational (IR), incentive compatible (IC), weakly budget balanced, and efficient. This has led follow-up work on two-sided trade settings to weaken the efficiency requirement and consider approximately efficient simple mechanisms, while still demanding the other properties. The current state-of-the-art of such mechanisms for two-sided markets can be categorized as giving one (but not both) of the following two types of approximation guarantees on the gains from trade: a constant ex-ante guarantee, measured with respect to the second-best efficiency benchmark, or an asymptotically optimal ex-post guarantee, measured with respect to the first-best efficiency benchmark. Here the second-best efficiency benchmark refers to the highest gains from trade attainable by any IR, IC and weakly budget balanced mechanism, while the first-best efficiency benchmark refers to the maximum gains from trade (attainable by the VCG mechanism, which is not weakly budget balanced). In this paper, we construct simple mechanisms for double-auction and matching markets that simultaneously achieve both types of guarantees: these are ex-post IR, Bayesian IC, and ex-post weakly budget balanced mechanisms that 1) ex-ante guarantee a constant fraction of the gains from trade of the second-best, and 2) ex-post guarantee a realization-dependent fraction of the gains from trade of the first-best, such that this realization-dependent fraction converges to 1 (full efficiency) as the market grows large.

cs.GT↗

arXiv2017

A Constructive Approach to Reduced-Form Auctions with Applications to Multi-Item Mechanism Design

We provide a constructive proof of Border's theorem [Bor91, HR15a] and its generalization to reduced-form auctions with asymmetric bidders [Bor07, MV10, CKM13]. Given a reduced form, we identify a subset of Border constraints that are necessary and sufficient to determine its feasibility. Importantly, the number of these constraints is linear in the total number of bidder types. In addition, we provide a characterization result showing that every feasible reduced form can be induced by an ex-post allocation rule that is a distribution over ironings of the same total ordering of the union of all bidders' types. We show how to leverage our results for single-item reduced forms to design auctions with heterogeneous items and asymmetric bidders with valuations that are additive over items. Appealing to our constructive Border's theorem, we obtain polynomial-time algorithms for computing the revenue-optimal auction. Appealing to our characterization of feasible reduced forms, we characterize feasible multi-item allocation rules.

cs.GT↗

arXiv2017

Learning Multi-item Auctions with (or without) Samples

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.

cs.GT↗

arXiv2017

Approximating Gains from Trade in Two-sided Markets via Simple Mechanisms

We design simple mechanisms to approximate the Gains from Trade (GFT) in two-sided markets with multiple unit-supply sellers and multiple unit-demand buyers. A classical impossibility result by Myerson and Satterthwaite showed that even with only one seller and one buyer, no Individually Rational (IR), Bayesian Incentive Compatible (BIC) and Budget-Balanced (BB) mechanism can achieve full GFT (trade whenever buyer's value is higher than the seller's cost). On the other hand, they proposed the "second-best" mechanism that maximizes the GFT subject to IR, BIC and BB constraints, which is unfortunately rather complex for even the single-seller single-buyer case. Our mechanism is simple, IR, BIC and BB, and achieves $\frac{1}{2}$ of the optimal GFT among all IR, BIC and BB mechanisms. Our result holds for arbitrary distributions of the buyers' and sellers' values and can accommodate any downward-closed feasibility constraints over the allocations. The analysis of our mechanism is facilitated by extending the Cai-Weinberg-Devanur duality framework to two-sided markets.

cs.GT↗

arXiv2015

Welfare and Rationality Guarantees for the Simultaneous Multiple-Round Ascending Auction

The simultaneous multiple-round auction (SMRA) and the combinatorial clock auction (CCA) are the two primary mechanisms used to sell bandwidth. Under truthful bidding, the SMRA is known to output a Walrasian equilibrium that maximizes social welfare provided the bidder valuation functions satisfy the gross substitutes property. Recently, it was shown that the combinatorial clock auction (CCA) provides good welfare guarantees for general classes of valuation functions. This motivates the question of whether similar welfare guarantees hold for the SMRA in the case of general valuation functions. We show the answer is no. But we prove that good welfare guarantees still arise if the degree of complementarities in the bidder valuations are bounded. In particular, if bidder valuations functions are $α$-near-submodular then, under truthful bidding, the SMRA has a welfare ratio (the worst case ratio between the social welfare of the optimal allocation and the auction allocation) of at most $(1+α)$. The special case of submodular valuations, namely $α=1$, and produces individually rational solutions. However, for $α>1$, this is a bicriteria guarantee, to obtain good welfare under truthful bidding requires relaxing individual rationality. Finally, we examine what strategies are required to ensure individual rationality in the SMRA with general valuation functions. First, we provide a weak characterization, namely \emph{secure bidding}, for individual rationality. We then show that if the bidders use a profit-maximizing secure bidding strategy the welfare ratio is at most $1+α$. Consequently, by bidding securely, it is possible to obtain the same welfare guarantees as truthful bidding without the loss of individual rationality.

cs.GT↗

arXiv2015

On the Economic Efficiency of the Combinatorial Clock Auction

Since the 1990s spectrum auctions have been implemented world-wide. This has provided for a practical examination of an assortment of auction mechanisms and, amongst these, two simultaneous ascending price auctions have proved to be extremely successful. These are the simultaneous multiround ascending auction (SMRA) and the combinatorial clock auction (CCA). It has long been known that, for certain classes of valuation functions, the SMRA provides good theoretical guarantees on social welfare. However, no such guarantees were known for the CCA. In this paper, we show that CCA does provide strong guarantees on social welfare provided the price increment and stopping rule are well-chosen. This is very surprising in that the choice of price increment has been used primarily to adjust auction duration and the stopping rule has attracted little attention. The main result is a polylogarithmic approximation guarantee for social welfare when the maximum number of items demanded $\mathcal{C}$ by a bidder is fixed. Specifically, we show that either the revenue of the CCA is at least an $Ω\Big(\frac{1}{\mathcal{C}^{2}\log n\log^2m}\Big)$-fraction of the optimal welfare or the welfare of the CCA is at least an $Ω\Big(\frac{1}{\log n}\Big)$-fraction of the optimal welfare, where $n$ is the number of bidders and $m$ is the number of items. As a corollary, the welfare ratio -- the worst case ratio between the social welfare of the optimum allocation and the social welfare of the CCA allocation -- is at most $O(\mathcal{C}^2 \cdot \log n \cdot \log^2 m)$. We emphasize that this latter result requires no assumption on bidders valuation functions. Finally, we prove that such a dependence on $\mathcal{C}$ is necessary. In particular, we show that the welfare ratio of the CCA is at least $Ω\Big(\mathcal{C} \cdot \frac{\log m}{\log \log m}\Big)$.

cs.GT↗

arXiv2015

Optimum Statistical Estimation with Strategic Data Sources

We propose an optimum mechanism for providing monetary incentives to the data sources of a statistical estimator such as linear regression, so that high quality data is provided at low cost, in the sense that the sum of payments and estimation error is minimized. The mechanism applies to a broad range of estimators, including linear and polynomial regression, kernel regression, and, under some additional assumptions, ridge regression. It also generalizes to several objectives, including minimizing estimation error subject to budget constraints. Besides our concrete results for regression problems, we contribute a mechanism design framework through which to design and analyze statistical estimators whose examples are supplied by workers with cost for labeling said examples.

stat.ML↗

arXiv2014

Extreme-Value Theorems for Optimal Multidimensional Pricing

We provide a near-optimal, computationally efficient algorithm for the unit-demand pricing problem, where a seller wants to price n items to optimize revenue against a unit-demand buyer whose values for the items are independently drawn from known distributions. For any chosen accuracy eps>0 and item values bounded in [0,1], our algorithm achieves revenue that is optimal up to an additive error of at most eps, in polynomial time. For values sampled from Monotone Hazard Rate (MHR) distributions, we achieve a (1-eps)-fraction of the optimal revenue in polynomial time, while for values sampled from regular distributions the same revenue guarantees are achieved in quasi-polynomial time. Our algorithm for bounded distributions applies probabilistic techniques to understand the statistical properties of revenue distributions, obtaining a reduction in the search space of the algorithm via dynamic programming. Adapting this approach to MHR and regular distributions requires the proof of novel extreme value theorems for such distributions. As a byproduct, our techniques establish structural properties of approximately-optimal and near-optimal solutions. We show that, for values independently distributed according to MHR distributions, pricing all items at the same price achieves a constant fraction of the optimal revenue. Moreover, for all eps >0, g(1/eps) distinct prices suffice to obtain a (1-eps)-fraction of the optimal revenue, where g(1/eps) is quadratic in 1/eps and independent of n. Similarly, for all eps>0 and n>0, at most g(1/(eps log n)) distinct prices suffice if the values are independently distributed according to regular distributions, where g() is a polynomial function. Finally, when the values are i.i.d. from some MHR distribution, we show that, if n is a sufficiently large function of 1/eps, a single price suffices to achieve a (1-eps)-fraction of the optimal revenue.

cs.GT↗

arXiv2013

Designing Markets for Daily Deals

Daily deals platforms such as Amazon Local, Google Offers, GroupOn, and LivingSocial have provided a new channel for merchants to directly market to consumers. In order to maximize consumer acquisition and retention, these platforms would like to offer deals that give good value to users. Currently, selecting such deals is done manually; however, the large number of submarkets and localities necessitates an automatic approach to selecting good deals and determining merchant payments. We approach this challenge as a market design problem. We postulate that merchants already have a good idea of the attractiveness of their deal to consumers as well as the amount they are willing to pay to offer their deal. The goal is to design an auction that maximizes a combination of the revenue of the auctioneer (platform), welfare of the bidders (merchants), and the positive externality on a third party (the consumer), despite the asymmetry of information about this consumer benefit. We design auctions that truthfully elicit this information from the merchants and maximize the social welfare objective, and we characterize the consumer welfare functions for which this objective is truthfully implementable. We generalize this characterization to a very broad mechanism-design setting and give examples of other applications.

cs.GT↗

arXiv2013

Reducing Revenue to Welfare Maximization: Approximation Algorithms and other Generalizations

It was recently shown in [http://arxiv.org/abs/1207.5518] that revenue optimization can be computationally efficiently reduced to welfare optimization in all multi-dimensional Bayesian auction problems with arbitrary (possibly combinatorial) feasibility constraints and independent additive bidders with arbitrary (possibly combinatorial) demand constraints. This reduction provides a poly-time solution to the optimal mechanism design problem in all auction settings where welfare optimization can be solved efficiently, but it is fragile to approximation and cannot provide solutions to settings where welfare maximization can only be tractably approximated. In this paper, we extend the reduction to accommodate approximation algorithms, providing an approximation preserving reduction from (truthful) revenue maximization to (not necessarily truthful) welfare maximization. The mechanisms output by our reduction choose allocations via black-box calls to welfare approximation on randomly selected inputs, thereby generalizing also our earlier structural results on optimal multi-dimensional mechanisms to approximately optimal mechanisms. Unlike [http://arxiv.org/abs/1207.5518], our results here are obtained through novel uses of the Ellipsoid algorithm and other optimization techniques over {\em non-convex regions}.

cs.GT↗

arXiv2013

Understanding Incentives: Mechanism Design becomes Algorithm Design

We provide a computationally efficient black-box reduction from mechanism design to algorithm design in very general settings. Specifically, we give an approximation-preserving reduction from truthfully maximizing \emph{any} objective under \emph{arbitrary} feasibility constraints with \emph{arbitrary} bidder types to (not necessarily truthfully) maximizing the same objective plus virtual welfare (under the same feasibility constraints). Our reduction is based on a fundamentally new approach: we describe a mechanism's behavior indirectly only in terms of the expected value it awards bidders for certain behavior, and never directly access the allocation rule at all. Applying our new approach to revenue, we exhibit settings where our reduction holds \emph{both ways}. That is, we also provide an approximation-sensitive reduction from (non-truthfully) maximizing virtual welfare to (truthfully) maximizing revenue, and therefore the two problems are computationally equivalent. With this equivalence in hand, we show that both problems are NP-hard to approximate within any polynomial factor, even for a single monotone submodular bidder. We further demonstrate the applicability of our reduction by providing a truthful mechanism maximizing fractional max-min fairness. This is the first instance of a truthful mechanism that optimizes a non-linear objective.

cs.GT↗

arXiv2012

Simple and Nearly Optimal Multi-Item Auctions

We provide a Polynomial Time Approximation Scheme (PTAS) for the Bayesian optimal multi-item multi-bidder auction problem under two conditions. First, bidders are independent, have additive valuations and are from the same population. Second, every bidder's value distributions of items are independent but not necessarily identical monotone hazard rate (MHR) distributions. For non-i.i.d. bidders, we also provide a PTAS when the number of bidders is small. Prior to our work, even for a single bidder, only constant factor approximations are known. Another appealing feature of our mechanism is the simple allocation rule. Indeed, the mechanism we use is either the second-price auction with reserve price on every item individually, or VCG allocation with a few outlying items that requires additional treatments. It is surprising that such simple allocation rules suffice to obtain nearly optimal revenue.

cs.GT↗