We consider the problem of equitably allocating a set of indivisible goods to n agents with additive utilities so as to provide worst case guarantees on agents' utilities. Demko and Hill [6] showed the existence of an allocation where every agent values his share at least V n (α), which is a family of nonincreasing functions of α, defined as the maximum value assigned by an agent to a single good. A deterministic algorithm returning such an allocation in polynomial time was proposed in [15]. Interestingly, V n (α) is tight for some values of α, i.e. it matches the highest possible utility of the least happy agent. However, this is not true for all values of α. We propose a family of functions W n such that W n (x) ≥ V n (x) for all x, and W n (x) > V n (x) for values of x where V n (x) is not tight. The functions W n apply on a problem that generalizes the allocation of indivisible goods. It is to find a base in a matroid which is common to n agents. Our results are constructive, they are achieved by analyzing an extension of the algorithm of Markakis and Psomas. We also present an upper bound on the utility of the least happy agent.

Abstract. We study a problem that generalizes the fair allocation of indivisible goods. The input is a matroid and a set of agents. Each agent has his own utility for every element of the matroid. Our goal is to build a base of the matroid and provide worst case guarantees on the additive utilities of the agents. These utilities are private, an assumption that is commonly made for the fair division of divisible resources, Since the use of an algorithm is not appropriate in this context, we resort to protocols, like in cake cutting problems. Our contribution is a protocol where the agents can interact and build a base of the matroid. If there are up to 8 agents, we show how everyone can ensure that his worst case utility for the resulting base is the same as those given by Markakis and Psomas [18] for the fair allocation of indivisible goods, based on the guarantees of Demko and Hill [8].

We consider problems where a solution is evaluated with a couple. Each coordinate of this couple represents the utility of an agent. Due to the possible conflicts, it is unlikely that one feasible solution is optimal for both agents. Then a natural aim is to find a tradeoff. We investigate tradeoff solutions with worst case guarantees for the agents. The focus is on discrete problems having a matroid structure and the utility of an agent is modeled with a function which is either additive or weighted labeled. We provide polynomial-time deterministic algorithms which achieve several guarantees and we prove that some guarantees are not possible to reach.

We introduce and study four optimization problems that generalize the well-known subset sum problem. Given a node-weighted digraph, select a subset of vertices whose total weight does not exceed a given budget. Some additional constraints need to be satisfied. The (weak resp.) digraph constraint imposes that if (all incoming nodes of resp.) a node x belongs to the solution, then the latter comprises all its outgoing nodes (node x itself resp.). The maximality constraint ensures that a solution cannot be extended without violating the budget or the (weak) digraph constraint. We study the complexity of these problems and we present some approximation results according to the type of digraph given in input, e.g. directed acyclic graphs and oriented trees.