Search

We study the House Allocation problem (also known as the Assignment problem), i.e., the problem of allocating a set of objects among a set of agents, where each agent has ordinal preferences (possibly involving ties) over a subset of the objects. We focus on truthful mechanisms without monetary transfers for finding large Pareto optimal matchings. It is straightforward to show that no deterministic truthful mechanism can approximate a maximum cardinality Pareto optimal matching with ratio better than 2. We thus consider randomised mechanisms. We give a natural and explicit extension of the classical Random Serial Dictatorship Mechanism (RSDM) specifically for the House Allocation problem where preference lists can include ties. We thus obtain a universally truthful randomised mechanism for finding a Pareto optimal matching and show that it achieves an approximation ratio of $\frac{e}{e-1}$. The same bound holds even when agents have priorities (weights) and our goal is to find a maximum weight (as opposed to maximum cardinality) Pareto optimal matching. On the other hand we give a lower bound of $\frac{18}{13}$ on the approximation ratio of any universally truthful Pareto optimal mechanism in settings with strict preferences. In the case that the mechanism must additionally be non-bossy with an additional technical assumption, we show by utilising a result of Bade that an improved lower bound of $\frac{e}{e-1}$ holds. This lower bound is tight since RSDM for strict preference lists is non-bossy. We moreover interpret our problem in terms of the classical secretary problem and prove that our mechanism provides the best randomised strategy of the administrator who interviews the applicants., To appear in Algorithmica (preliminary version appeared in the Proceedings of EC 2014)

Several countries successfully use centralized matching schemes for assigning students to study places or fresh graduates to their first positions. In this paper we explore the computational aspects of a possible similar scheme for assigning trainee teachers to schools. Our model is motivated by the situation characteristic for Slovak and Czech education system where each pre-service teacher specializes in two subjects. We show that if the two subjects can be performed independently in two different schools, then a feasible assignment can be found efficiently by employing network flow techniques. By contrast, the requirement to perform both subjects at the same school leads to intractable problems even under several strict restrictions concerning the total number of subjects, partial capacities of schools and the number of acceptable schools each teacher is allowed to list. Finally, we report on an integer programming model for solving the ‘inseparable subjects’ case of the teachers assignment problem and the results of its application to real data.

Matching problems with preferences occur in widespread applications such as the assignment of school-leavers to universities, junior doctors to hospitals, students to campus housing, children to schools, kidney transplant patients to donors and so on. The common thread is that agents have preference lists over the possible outcomes and the task is to find a matching of the participants that is in some sense optimal with respect to these preferences. The archetypal problem of this kind is the celebrated Stable Marriage problem, first introduced by David Gale and Lloyd Shapley in 1962. Over the last decade there has been something of a resurgence of interest in the algorithmic study of variants of this problem. In addition, research has been instigated into related matching problems, for example bipartite models with one-sided preferences, under criteria such as rank-maximality, popularity and Pareto optimality. These developments can be attributed in part to the growing tendency for matching processes to involve the online collection of preference information, necessitating efficient algorithms for constructing matchings that are fair and optimal in an expanding range of problem variants.

The Hospitals/Residents problem with Couples (HRC) is a generalisation of the classical Hospitals/Residents problem (HR) that is important in practical applications because it models the case where couples submit joint preference lists over pairs of hospitals (h i ,h j ). We consider a natural restriction of HRC in which the members of a couple have individual preference lists over hospitals, and the joint preference list of the couple is consistent with these individual lists in a precise sense. We give an appropriate stability definition and show that, in this context, the problem of deciding whether a stable matching exists is NP-complete, even if each resident’s preference list has length at most 3 and each hospital has capacity at most 2. However, with respect to classical (Gale-Shapley) stability, we give a linear-time algorithm to find a stable matching or report that none exists, regardless of the preference list lengths or the hospital capacities. Finally, for an alternative formulation of our restriction of HRC, which we call the Hospitals/Residents problem with Sizes (HRS), we give a linear-time algorithm that always finds a stable matching for the case that hospital preference lists are of length at most 2, and where hospital capacities can be arbitrary.

When ties and incomplete preference lists are permitted in the Stable Marriage and Hospitals/Residents problems, stable matchings can have different sizes. The problem of finding a maximum cardinality stable matching in this context is known to be NP-hard, even under very severe restrictions on the number, size and position of ties. In this paper, we describe polynomial-time 5/3-approximation algorithms for variants of these problems in which ties are on one side only and at the end of the preference lists. The particular variant is motivated by important applications in large scale centralised matching schemes.