arXiv subjects

Richard Cole

Publications and source records attributed to Richard Cole.

At least 19 recordsLinked to original sources

arXiv2025

Distributed Interview Selection for Stable Matching in Large Random Markets

In real-world settings of the Deferred Acceptance stable matching algorithm, such as the American medical residency match (NRMP), school choice programs, and various national university entrance systems, candidates need to decide which programs to list. In many of these settings there is an initial phase of interviews or information gathering which affect the preferences on one or both sides. We ask: which interviews should candidates seek? We study this question in a model, introduced by Lee (2016) and modified by Allman and Ashlagi (2023), with preferences based on correlated cardinal utilities. We describe a distributed, low-communication strategy for the doctors and students, which lead to non-match rates of $e^{(-\widetilde{O}(\sqrt{k}))}$ in the residency setting and $e^{(-\widetilde{O}(k))}$ in the school-choice setting, where $k$ is the number of interviews per doctor in the first setting, and the number of proposals per student in the second setting; these bounds do not apply to the agents with the lowest public ratings, the bottommost agents, who may not fare as well. We also obtain bounds on the expected utilities each non-bottommost agent obtains. These results are parameterized by the capacity of the hospital programs and schools. Larger capacities improve the outcome for the hospitals and schools, but don't significantly affect the outcomes of the doctors or students. Finally, in the school choice setting we obtain an $\epsilon$-Nash type equilibrium for the students apart from the bottommost ones; importantly, the equilibrium holds regardless of the actions of the bottommost students. We also discuss to what extent this result extends to the residency setting. We complement our theoretical results with an experimental study that shows the asymptotic results hold for real-world values of $n$.

cs.GT↗

arXiv2025

Proportional Response Dynamics in Gross Substitutes Markets

Proportional response is a well-established distributed algorithm which has been shown to converge to competitive equilibria in both Fisher and Arrow-Debreu markets, for various sub-families of homogeneous utilities, including linear and constant elasticity of substitution utilities. We propose a natural generalization of proportional response for gross substitutes utilities, and prove that it converges to competitive equilibria in Fisher markets. This is the first convergence result of a proportional response style dynamics in Fisher markets for utilities beyond the homogeneous utilities covered by the Eisenberg-Gale convex program. We show an empirical convergence rate of $O(1/T)$ for the prices. Furthermore, we show that the allocations of a lazy version of the generalized proportional response dynamics converge to competitive equilibria in Arrow-Debreu markets.

cs.GT↗

arXiv2025

Facility Location with Public Locations and Private Doubly-Peaked Costs

In the facility location problem, the task is to place one or more facilities so as to minimize the sum of the agent costs for accessing their nearest facility. Heretofore, in the strategic version, agent locations have been assumed to be private, while their cost measures have been public and identical. For the most part, the cost measure has been the distance to the nearest facility. However, in multiple natural settings, such as placing a firehouse or a school, this modeling does not appear to be a good fit. For it seems natural that the agent locations would be known, but their costs might be private information. In addition, for these types of settings, agents may well want the nearest facility to be at the right distance: near, but not too near. This is captured by the doubly-peaked cost introduced by Filos-Ratsikas et al. (AAMAS 2017). In this paper, we re-examine the facility location problem from this perspective: known agent locations and private preferred distances to the nearest facility. We then give lower and upper bounds on achievable approximations, focusing on the problem in 1D, and in 2D with an $L_1$ distance measure.

cs.GT↗

arXiv2023

A First Order Method for Linear Programming Parameterized by Circuit Imbalance

Various first order approaches have been proposed in the literature to solve Linear Programming (LP) problems, recently leading to practically efficient solvers for large-scale LPs. From a theoretical perspective, linear convergence rates have been established for first order LP algorithms, despite the fact that the underlying formulations are not strongly convex. However, the convergence rate typically depends on the Hoffman constant of a large matrix that contains the constraint matrix, as well as the right hand side, cost, and capacity vectors. We introduce a first order approach for LP optimization with a convergence rate depending polynomially on the circuit imbalance measure, which is a geometric parameter of the constraint matrix, and depending logarithmically on the right hand side, capacity, and cost vectors. This provides much stronger convergence guarantees. For example, if the constraint matrix is totally unimodular, we obtain polynomial-time algorithms, whereas the convergence guarantees for approaches based on primal-dual formulations may have arbitrarily slow convergence rates for this class. Our approach is based on a fast gradient method due to Necoara, Nesterov, and Glineur (Math. Prog. 2019); this algorithm is called repeatedly in a framework that gradually fixes variables to the boundary. This technique is based on a new approximate version of Tardos's method, that was used to obtain a strongly polynomial algorithm for combinatorial LPs (Oper. Res. 1986).

math.OC↗

arXiv2022

Stable Matching: Choosing Which Proposals to Make

To guarantee all agents are matched in general, the classic Deferred Acceptance algorithm needs complete preference lists. In practice, preference lists are short, yet stable matching still works well. This raises two questions: $\bullet$ Why does it work well? $\bullet$ Which proposals should agents include in their preference lists? We study these questions in a model, introduced by Lee [17], with preferences based on correlated cardinal utilities: these utilities are based on common public ratings of each agent together with individual private adjustments. Lee showed that for suitable utility functions, in large markets, with high probability, for most agents, all stable matchings yield similar valued utilities. By means of a new analysis, we strengthen Lee's result, showing that in large markets, with high probability, for $\it all$ but the agents with the lowest public ratings, all stable matchings yield similar valued utilities. We can then deduce that for $\it all$ but the agents with the lowest public ratings, each agent has an easily identified length $O(\log n)$ preference list that includes all of its stable matches, addressing the second question above. We note that this identification uses an initial communication phase. We extend these results to settings where the two sides have unequal numbers of agents, to many-to-one settings, e.g. employers and workers, and we also show the existence of an $\epsilon$-Bayes-Nash equilibrium in which every agent makes relatively few proposals. These results all rely on a new technique for sidestepping the conditioning between the tentative matching events that occur over the course of a run of the Deferred Acceptance algorithm. We complement these theoretical results with an experimental study.

cs.GT↗

arXiv2021

Selecting a Match: Exploration vs Decision

In a dynamic matching market, such as a marriage or job market, how should agents balance accepting a proposed match with the cost of continuing their search? We consider this problem in a discrete setting, in which agents have cardinal values and finite lifetimes, and proposed matches are random. We seek to quantify how well the agents can do. We provide upper and lower bounds on the collective losses of the agents, with a polynomially small failure probability, where the notion of loss is with respect to a plausible baseline we define. These bounds are tight up to constant factors. We highlight two aspects of this work. First, in our model, agents have a finite time in which to enjoy their matches, namely the minimum of their remaining lifetime and that of their partner; this implies that unmatched agents become less desirable over time, and suggests that their decision rules should change over time. Second, we use a discrete rather than a continuum model for the population. The discreteness causes variance which induces localized imbalances in the two sides of the market. One of the main technical challenges we face is to bound these imbalances. In addition, we present the results of simulations on moderate-sized problems for both the discrete and continuum versions. For these size problems, there are substantial ongoing fluctuations in the discrete setting whereas the continuum version converges reasonably quickly.

cs.GT↗

arXiv2020

Non-quasi-linear Agents in Quasi-linear Mechanisms

Mechanisms with money are commonly designed under the assumption that agents are quasi-linear, meaning they have linear disutility for spending money. We study the implications when agents with non-linear (specifically, convex) disutility for payments participate in mechanisms designed for quasi-linear agents. We first show that any mechanism that is truthful for quasi-linear buyers has a simple best response function for buyers with non-linear disutility from payments, in which each bidder simply scales down her value for each potential outcome by a fixed factor, equal to her target return on investment (ROI). We call such a strategy ROI-optimal. We prove the existence of a Nash equilibrium in which agents use ROI-optimal strategies for a general class of allocation problems. Motivated by online marketplaces, we then focus on simultaneous second-price auctions for additive bidders and show that all ROI-optimal equilibria in this setting achieve constant-factor approximations to suitable welfare and revenue benchmarks.

cs.GT↗

arXiv2020

Parallel Stochastic Asynchronous Coordinate Descent: Tight Bounds on the Possible Parallelism

Several works have shown linear speedup is achieved by an asynchronous parallel implementation of stochastic coordinate descent so long as there is not too much parallelism. More specifically, it is known that if all updates are of similar duration, then linear speedup is possible with up to $Θ(\sqrt n L_{\max}/L_{\overline{\mathrm{res}}})$ processors, where $L_{\max}$ and $L_{\overline{\mathrm{res}}}$ are suitable Lipschitz parameters. This paper shows the bound is tight for almost all possible values of these parameters.

math.OC↗

arXiv2020

Fully Asynchronous Stochastic Coordinate Descent: A Tight Lower Bound on the Parallelism Achieving Linear Speedup

We seek tight bounds on the viable parallelism in asynchronous implementations of coordinate descent that achieves linear speedup. We focus on asynchronous coordinate descent (ACD) algorithms on convex functions which consist of the sum of a smooth convex part and a possibly non-smooth separable convex part. We quantify the shortfall in progress compared to the standard sequential stochastic gradient descent. This leads to a simple yet tight analysis of the standard stochastic ACD in a partially asynchronous environment, generalizing and improving the bounds in prior work. We also give a considerably more involved analysis for general asynchronous environments in which the only constraint is that each update can overlap with at most q others. The new lower bound on the maximum degree of parallelism attaining linear speedup is tight and improves the best prior bound almost quadratically.

math.OC↗

arXiv2020

A Truthful Cardinal Mechanism for One-Sided Matching

We revisit the well-studied problem of designing mechanisms for one-sided matching markets, where a set of $n$ agents needs to be matched to a set of $n$ heterogeneous items. Each agent $i$ has a value $v_{i,j}$ for each item $j$, and these values are private information that the agents may misreport if doing so leads to a preferred outcome. Ensuring that the agents have no incentive to misreport requires a careful design of the matching mechanism, and mechanisms proposed in the literature mitigate this issue by eliciting only the \emph{ordinal} preferences of the agents, i.e., their ranking of the items from most to least preferred. However, the efficiency guarantees of these mechanisms are based only on weak measures that are oblivious to the underlying values. In this paper we achieve stronger performance guarantees by introducing a mechanism that truthfully elicits the full \emph{cardinal} preferences of the agents, i.e., all of the $v_{i,j}$ values. We evaluate the performance of this mechanism using the much more demanding Nash bargaining solution as a benchmark, and we prove that our mechanism significantly outperforms all ordinal mechanisms (even non-truthful ones). To prove our approximation bounds, we also study the population monotonicity of the Nash bargaining solution in the context of matching markets, providing both upper and lower bounds which are of independent interest.

cs.GT↗

arXiv2019

Balancing the Robustness and Convergence of Tatonnement

A major goal in Algorithmic Game Theory is to justify equilibrium concepts from an algorithmic and complexity perspective. One appealing approach is to identify robust natural distributed algorithms that converge quickly to an equilibrium. This paper addresses a lack of robustness in existing convergence results for discrete forms of tatonnement, including the fact that it need not converge when buyers have linear utility functions. This work achieves greater robustness by seeking approximate rather than exact convergence in large market settings. More specifically, this paper shows that for Fisher markets with buyers having CES utility functions, including linear utility functions, tatonnement will converge quickly to an approximate equilibrium (i.e. at a linear rate), modulo a suitable large market assumption. The quality of the approximation is a function of the parameters of the large market assumption.

cs.GT↗

arXiv2019

On the Existence of Pareto Efficient and Envy Free Allocations

Envy-freeness and Pareto Efficiency are two major goals in welfare economics. The existence of an allocation that satisfies both conditions has been studied for a long time. Whether items are indivisible or divisible, it is impossible to achieve envy-freeness and Pareto Efficiency ex post even in the case of two people and two items. In contrast, in this work, we prove that, for any cardinal utility functions (including complementary utilities for example) and for any number of items and players, there always exists an ex ante mixed allocation which is envy-free and Pareto Efficient, assuming the allowable assignments are closed under swaps, i.e. if given a legal assignment, swapping any two players' allocations produces another legal assignment. The problem remains open in the divisible case. We also investigate the communication complexity for finding a Pareto Efficient and envy-free allocation.

cs.GT↗

arXiv2018

An Analysis of Asynchronous Stochastic Accelerated Coordinate Descent

Gradient descent, and coordinate descent in particular, are core tools in machine learning and elsewhere. Large problem instances are common. To help solve them, two orthogonal approaches are known: acceleration and parallelism. In this work, we ask whether they can be used simultaneously. The answer is "yes". More specifically, we consider an asynchronous parallel version of the accelerated coordinate descent algorithm proposed and analyzed by Lin, Liu and Xiao (SIOPT'15). We give an analysis based on the efficient implementation of this algorithm. The only constraint is a standard bounded asynchrony assumption, namely that each update can overlap with at most q others. (q is at most the number of processors times the ratio in the lengths of the longest and shortest updates.) We obtain the following three results: 1. A linear speedup for strongly convex functions so long as q is not too large. 2. A substantial, albeit sublinear, speedup for strongly convex functions for larger q. 3. A substantial, albeit sublinear, speedup for convex functions.

math.OC↗

arXiv2018

When Does Diversity of User Preferences Improve Outcomes in Selfish Routing?

We seek to understand when heterogeneity in user preferences yields improved outcomes in terms of overall cost. That this might be hoped for is based on the common belief that diversity is advantageous in many settings. We investigate this in the context of routing. Our main result is a sharp characterization of the network settings in which diversity always helps, versus those in which it is sometimes harmful. Specifically, we consider routing games, where diversity arises in the way that users trade-off two criteria (such as time and money, or, in the case of stochastic delays, expectation and variance of delay). Our main contributions are the following: 1) A participant-oriented measure of cost in the presence of user diversity, together with the identification of the natural benchmark: the same cost measure for an appropriately defined average of the diversity. 2) A full characterization of those network topologies for which diversity always helps, for all latency functions and demands. For single-commodity routings, these are series-parallel graphs, while for multi-commodity routings, they are the newly-defined "block-matching" networks. The latter comprise a suitable interweaving of multiple series-parallel graphs each connecting a distinct source-sink pair. While the result for the single-commodity case may seem intuitive in light of the well-known Braess paradox, the two problems are different: there are instances where diversity helps although the Braess paradox occurs, and vice-versa. But the main technical challenge is to establish the "only if" direction of the result for multi-commodity networks. This follows by constructing an instance where diversity hurts, and showing how to embed it in any network which is not block-matching, by carefully exploiting the way the simple source-sink paths of the commodities intersect in the "non-block-matching" portion of the network.

cs.GT↗

arXiv2018

Amortized Analysis of Asynchronous Price Dynamics

We extend a recently developed framework for analyzing asynchronous coordinate descent algorithms to show that an asynchronous version of tatonnement, a fundamental price dynamic widely studied in general equilibrium theory, converges toward a market equilibrium for Fisher markets with CES utilities or Leontief utilities, for which tatonnement is equivalent to coordinate descent.

cs.GT↗

arXiv2018

Dynamics of Distributed Updating in Fisher Markets

A major goal in Algorithmic Game Theory is to justify equilibrium concepts from an algorithmic and complexity perspective. One appealing approach is to identify natural distributed algorithms that converge quickly to an equilibrium. This paper established new convergence results for two generalizations of Proportional Response in Fisher markets with buyers having CES utility functions. The starting points are respectively a new convex and a new convex-concave formulation of such markets. The two generalizations correspond to suitable mirror descent algorithms applied to these formulations. Several of our new results are a consequence of new notions of strong Bregman convexity and of strong Bregman convex-concave functions, and associated linear rates of convergence, which may be of independent interest. Among other results, we analyze a damped generalized Proportional Response and show a linear rate of convergence in a Fisher market with buyers whose utility functions cover the full spectrum of CES utilities aside the extremes of linear and Leontief utilities; when these utilities are included, we obtain an empirical O(1/T) rate of convergence.

cs.GT↗

arXiv2017

Large Market Games with Near Optimal Efficiency

As is well known, many classes of markets have efficient equilibria, but this depends on agents being non-strategic, i.e. that they declare their true demands when offered goods at particular prices, or in other words, that they are price-takers. An important question is how much the equilibria degrade in the face of strategic behavior, i.e. what is the Price of Anarchy (PoA) of the market viewed as a mechanism? Often, PoA bounds are modest constants such as 4/3 or 2. Nonetheless, in practice a guarantee that no more than 25% or 50% of the economic value is lost may be unappealing. This paper asks whether significantly better bounds are possible under plausible assumptions. In particular, we look at how these worst case guarantees improve in the following large settings. Large Walrasian auctions: These are auctions with many copies of each item and many agents. We show that the PoA tends to 1 as the market size increases, under suitable conditions, mainly that there is some uncertainty about the numbers of copies of each good and demands obey the gross substitutes condition. We also note that some such assumption is unavoidable. Large Fisher markets: Fisher markets are a class of economies that has received considerable attention in the computer science literature. A large market is one in which at equilibrium, each buyer makes only a small fraction of the total purchases; the smaller the fraction, the larger the market. Here the main condition is that demands are based on homogeneous monotone utility functions that satisfy the gross substitutes condition. Again, the PoA tends to 1 as the market size increases. Furthermore, in each setting, we quantify the tradeoff between market size and the PoA.

cs.GT↗

arXiv2017

Resource Oblivious Sorting on Multicores

We present a deterministic sorting algorithm, SPMS (Sample, Partition, and Merge Sort), that interleaves the partitioning of a sample sort with merging. Sequentially, it sorts $n$ elements in $O(n \log n)$ time cache-obliviously with an optimal number of cache misses. The parallel complexity (or critical path length) of the algorithm is $O(\log n \cdot \log\log n)$, which improves on previous bounds for optimal cache oblivious sorting. The algorithm also has low false sharing costs. When scheduled by a work-stealing scheduler in a multicore computing environment with a global shared memory and $p$ cores, each having a cache of size $M$ organized in blocks of size $B$, the costs of the additional cache misses and false sharing misses due to this parallel execution are bounded by the cost of $O(S\cdot M/B)$ and $O(S \cdot B)$ cache misses respectively, where $S$ is the number of steals performed during the execution. Finally, SPMS is resource oblivious in Athat the dependence on machine parameters appear only in the analysis of its performance, and not within the algorithm itself.

cs.DS↗