From convex optimization to randomized mechanisms: toward optimal combinatorial auctions

@article{Dughmi2011FromCO,
  title={From convex optimization to randomized mechanisms: toward optimal combinatorial auctions},
  author={Shaddin Dughmi and Tim Roughgarden and Qiqi Yan},
  journal={ArXiv},
  year={2011},
  volume={abs/1103.0040}
}
We design an expected polynomial time, truthful in expectation, (1-1/e)-approximation mechanism for welfare maximization in a fundamental class of combinatorial auctions. Our results apply to bidders with valuations that are matroid rank sums (MRS), which encompass most concrete examples of submodular functions studied in this context, including coverage functions and matroid weighted-rank functions. Our approximation factor is the best possible, even for known and explicitly given coverage… 
Optimal Mechanisms for Combinatorial Auctions and Combinatorial Public Projects via Convex Rounding
We design the first truthful-in-expectation, constant-factor approximation mechanisms for NP-hard cases of the welfare maximization problem in combinatorial auctions with nonidentical items and in
Limitations of Randomized Mechanisms for Combinatorial Auctions
  • S. Dughmi, J. Vondrák
  • Computer Science
    2011 IEEE 52nd Annual Symposium on Foundations of Computer Science
  • 2011
TLDR
It is proved that there is a constant \gamma>0 such that there are no truthful-in-expectation mechanisms that achieves an m-\gamma-approximation to the optimal social welfare for combinatorial public projects with sub modular valuations in the value oracle model.
Improved Truthful Mechanisms for Combinatorial Auctions with Submodular Bidders
TLDR
This work presents a computationally-efficient truthful mechanism with approximation ratio that improves upon the state-of-the-art by an exponential factor and achieves an O((log logm)^3) -approximation in expectation, uses only O(n) demand queries, and has universal truthfulness guarantee.
A Truthful-in-Expectation Mechanism for the Generalized Assignment Problem
TLDR
This work presents a novel convex optimization program for the auction which results in a maximal-in-distributional-range (MIDR) allocation rule and simplifies the existing optimization algorithms for the GAP while the approximation ratio is comparable to the best given approximation.
Breaking the logarithmic barrier for truthful combinatorial auctions with submodular bidders
TLDR
This paper develops the first mechanism that breaks the logarithmic barrier, and provides an approximation ratio of O( m).
Impossibility Results for Truthful Combinatorial Auctions with Submodular Valuations
TLDR
It is shown that every universally truthful randomized mechanism for combinatorial auctions with submodular valuations that approximates optimal social welfare within a factor of m1/2−ε must use exponentially many value queries, where m is the number of items.
From query complexity to computational complexity
TLDR
The technique relies instead on a direct reduction from Unique-SAT using list-decodable codes, and provides a general way of translating oracle inapproximability results arising from the symmetry gap technique to computational complexity inapp approximation results, where the submodular function is given explicitly.
A Truthful Mechanism for the Generalized Assignment Problem
TLDR
A truthful-in-expectation, (1-1/e)-approximation mechanism for a strategic variant of the generalized assignment problem (GAP), which improves upon the existing optimization algorithms for GAP in terms of simplicity and runtime while the approximation ratio closely matches the best approximation ratio known for G AP when all inputs are publicly known.
On the hardness of dominant strategy mechanism design
TLDR
This work proves the first gap between computationally efficient deterministic dominant strategy mechanisms and randomized ones, and also solves some open problems in the area of simultaneous combinatorial auctions.
Truthful Mechanism Design via Correlated Tree Rounding
TLDR
A simple and natural correlated rounding technique for designing mechanisms that are truthful in expectation and gives a truthful 2-approximate MIDR mechanism without using the ellipsoid method or convex programming.
...
...

References

SHOWING 1-10 OF 34 REFERENCES
Tight information-theoretic lower bounds for welfare maximization in combinatorial auctions
TLDR
Tight information-theoretic lower bounds are provided for the welfare maximization problem in combinatorial auctions and the well-known value query model in which the permitted query to a valuation function is in the form of a subset of items, and the reply is the value assigned to that subset of Items by the valuation function.
Truth revelation in approximately efficient combinatorial auctions
TLDR
It is shown that the GVA payment scheme does not provide for a truth revealing mechanism, and another scheme is introduced that does guarantee truthfulness for a restricted class of players.
Truthful and near-optimal mechanism design via linear programming
  • R. Lavi, Chaitanya Swamy
  • Computer Science
    46th Annual IEEE Symposium on Foundations of Computer Science (FOCS'05)
  • 2005
TLDR
The construction of the first truthful mechanisms with approximation guarantees for a variety of multi-parameter domains can be seen as a way of exploiting VCG in a computational tractable way even when the underlying social-welfare maximization problem is NP-hard.
An improved approximation algorithm for combinatorial auctions with submodular bidders
TLDR
By presenting a matching unconditional lower bound in the communication model, it is proved that the upper bound is tight for the XOS class, and the algorithm improves upon the previously known 2-approximation algorithm.
Inapproximability Results for Combinatorial Auctions with Submodular Utility Functions
TLDR
It is proved that there is no polynomial time approximation algorithm which approximates the maximum social welfare by a factor better than 1–1/e ≃ 0.632, unless P = NP.
Towards a characterization of truthful combinatorial auctions
TLDR
It is shown that truthful mechanisms for combinatorial auctions (and related restricted domains) must be "almost affine maximizers" if they also satisfy an additional requirement of "independence of irrelevant alternatives" and this implies the near-helplessness of such truthful polynomial-time auctions in all cases where exact optimization is computationally intractable.
Two Randomized Mechanisms for Combinatorial Auctions
TLDR
A new framework for developing truthful randomized mechanisms enables the construction of mechanisms with polynomially small failure probability, and provides the best approximation ratio for combinatorial auctions with submodularbidders currently achieved by truthful mechanisms.
On maximizing welfare when utility functions are subadditive
  • U. Feige
  • Economics, Mathematics
    STOC '06
  • 2006
TLDR
A way of rounding any fractional solution to a linear programming relaxation to solve the problem of maximizing welfare so as to give a feasible solution of welfare at least half that of the value of the fractional Solution.
Optimal approximation for the submodular welfare problem in the value oracle model
TLDR
A randomized continuous greedy algorithm is developed which achieves a (1-1/e)-approximation for the Submodular Welfare Problem in the value oracle model and is shown to have a potential of wider applicability on the examples of the Generalized Assignment Problem and the AdWords Assignment Problem.
Approximation algorithms for allocation problems: Improving the factor of 1 - 1/e
  • U. Feige, J. Vondrák
  • Mathematics
    2006 47th Annual IEEE Symposium on Foundations of Computer Science (FOCS'06)
  • 2006
TLDR
The 1 - 1/e approximation ratio is shown, and APX-hardness results for SMW and GAP are proved, which show that they apply even in very restricted settings, e.g. when every player has nonzero utility only for a constant number of items.
...
...