We study a general framework of optimization with the aim to compute fair solutions in settings with a set of agents whose valuations are combined using an aggregation function. The strength of our framework lies (1) in its generality and (2) in the fact that we leverage the power of ex-ante fairness, a concept that has recently gained much attention in the scope of fair allocation and fairness in AI in general. More precisely, in our setting there are n set functions f₁, …, fₙ (e.g., the valuation functions of n agents) that are combined using an aggregation function g (e.g., the minimum, Nash social welfare, p-norm). The power of ex-ante fairness is obtained by allowing as a feasible solution not simply a finite set S, but instead a distribution Π over feasible sets. The goal in our setting is then to find a probability distribution p in Π that maximizes the value resulting from aggregating (using g) the n expected values of the functions f₁, …, fₙ obtained when sampling a set S according to the distribution p. We stress that this is different from maximizing the expected value of g (ex-post fairness) and typically allows for much fairer solutions. We give three different greedy algorithms for three different settings of this framework and prove that they achieve constant approximation guarantees under certain realistic assumptions. For some of the settings, we show that these approximation guarantees are tight. Specific scenarios that can be modelled using our framework include fair information diffusion in social networks, fair submodular matching problems, and ex-ante versions of item assignment problems.

We revisit the setting of fair allocation of indivisible items among agents with heterogeneous, non-monotone valuations. We explore the existence and efficient computation of allocations that approximately satisfy either envy-freeness or equity constraints. Approximate envy-freeness ensures that each agent values her bundle at least as much as those given to the others, after some (or any) item removal, while approximate equity guarantees roughly equal valuations among agents, under similar adjustments. As a key technical contribution of this work, by leveraging fixed-point theorems (such as Sperner's Lemma and its variants), we establish the existence of envy-free-up-to-one-good-and-one-chore (EF1_g^c) and equitable-up-to-one-good-and-one-chore (EQ1_g^c) allocations, for non-monotone valuations that are always either non-negative or non-positive. These notions represent slight relaxations of the well-studied envy-free-up-to-one-item (EF1) and equitable-up-to-one-item (EQ1) guarantees, respectively. Our existential results hold even when items are arranged in a path and bundles must form connected sub-paths. The case of non-positive valuations, in particular, has been solved by proving a novel multi-coloring variant of Sperner's Lemma that constitutes a combinatorial result of independent interest. In addition, we also design a polynomial-time dynamic programming algorithm that computes an EQ1_g^c allocation. For monotone non-increasing valuations and path-connected bundles, all the above results can be extended to EF1 and EQ1 guarantees as well. Finally, we provide existential and computational results for certain stronger up-to-any-item equity notions under objective valuations, where items are partitioned into goods and chores.

In this work, we consider the problem of improving the efficiency of utility-sharing games, by resorting to a limited amount of subsidies. Utility-sharing games model scenarios in which strategic and self-interested players interact with each other by selecting resources. Each resource produces a utility that depends on the number of players selecting it, as a non-negative, non-decreasing and concave function, and each of these players receives an equal share of this utility. As the players’ selfish behavior may lead to pure Nash equilibria whose total utility is sub-optimal, previous work has resorted to subsidies, incentivizing the use of some resources, to contrast this phenomenon. We focus on the case in which the budget used to provide subsidies is bounded. We consider a class of mechanisms, called α -subsidy mechanisms, that allocate the budget in such a way that each player’s payoff is re-scaled up to a factor α ≥ 1. We design a specific sub-class of α -subsidy mechanisms, that can be implemented efficiently and distributedly by each resource, and evaluate their efficiency by providing upper bounds on their price of anarchy. These bounds are parametrized by both α and the underlying utility functions and are shown to be best-possible for α -subsidy mechanisms. Finally, we apply our results to the particular case of monomial utility functions of degree p ∈ (0, 1), and derive bounds on the price of anarchy that are parametrized by p and α .

Respiratory diseases are among the leading causes of morbidity worldwide, making timely and accurate diagnosis essential. However, interpreting chest X-rays is challenging due to the variability of pathological manifestations and the subjectivity of human analysis. In this study, we propose a multimodal approach that integrates automated image analysis with textual clinical data, leveraging a Visual Question Answering (VQA)based architecture and a text generation model for diagnostic report production. The use of Grad-CAM enhances the interpretability of the system by highlighting the most relevant image regions for diagnosis. The model was trained on a balanced dataset obtained by merging three sources-Lung X-ray Data, NIH Chest X-rays, and Chest X-Ray Images-ensuring fair classification across Normal and Pneumonia categories. The pipeline includes visual feature extraction using a Vision Transformer (ViT), automatic pathology classification, and diagnostic report generation with an advanced language model. Results indicate a significant improvement in diagnostic accuracy compared to traditional methods, supported by key performance metrics such as accuracy 95.3%, sensitivity, specificity, and F1-score. Furthermore, integrating the system into an interactive web app facilitates clinical adoption, enhancing diagnostic efficiency and supporting personalized management of pulmonary diseases.

We address the problem of improving the worst-case efficiency of pure Nash equilibria (aka, the price of anarchy) in affine congestion games, through a novel use of signalling. We assume that, for each player in the game, a most preferred strategy is publicly signalled. This can be done either distributedly by the players themselves, or be the outcome of some centralized algorithm. We apply this signalling scheme to two well-studied scenarios: games with partially altruistic players and games with resource taxation. We show a significant improvement in the price of anarchy of these games, whenever the aggregate signalled strategy profile is a good approximation of the game social optimum.

We reconsider atomic and non-atomic affine congestion games under the assumption that players are partitioned into p priority classes and resources schedule their users according to a priority-based policy, breaking ties uniformly at random. We derive tight bounds on both the price of anarchy and the price of stability as a function of p, revealing an interesting separation between the general case of p≥2 and the priority-free scenario of p=1. In fact, while in absence of priorities the worst-case prices of anarchy and stability of non-atomic games are lower than their counterparts in atomic ones, the two classes share the same bounds when p≥2. Moreover, while the worst-case price of stability is lower than the worst-case price of anarchy in atomic games with no priorities, their values become equal when p≥2. Said differently, the presence of priorities simultaneously irons out any combinatorial difference between atomic and non-atomic requests and among different pure Nash equilibria to produce a unique representative worst-case situation. Notably, our results keep holding even under singleton strategies. Besides being of independent interest, priority-based scheduling shares tight connections with online load balancing and finds a natural application within the theory of coordination mechanisms and cost-sharing policies for congestion games. Under this perspective, a number of possible research directions also arise.

Modeling how agents form their opinions is of paramount importance for designing marketing and electoral campaigns. In this work, we present a new framework for opinion formation which generalizes the well-known Friedkin-Johnsen model by incorporating three important features: (i) social group membership, that limits the amount of influence that people not belonging to the same group may lead on a given agent; (ii) both attraction among friends, and repulsion among enemies; (iii) different strengths of influence lead from different people on a given agent, even if the social relationships among them are the same. We show that, despite its generality, our model always admits a pure Nash equilibrium which, under opportune mild conditions, is even unique. Next, we analyze the performances of these equilibria with respect to a social objective function defined as a convex combination, parametrized by a value λ∈[0,1], of the costs yielded by the untruthfulness of the declared opinions and the total cost of social pressure. We prove bounds on both the price of anarchy and the price of stability which show that, for not-too-extreme values of λ, performance at equilibrium are very close to optimal ones. For instance, in several interesting scenarios, the prices of anarchy and stability are both equal to max{2λ,1-λ}/min{2λ,1-λ} which never exceeds 2 for λ∈[1/5,1/2].

In the influence maximization (IM) problem, we are given a social network and a budget k, and we look for a set of k nodes in the network, called seeds, that maximize the expected number of nodes that are reached by an influence cascade generated by the seeds, according to some stochastic model for influence diffusion. Extensive studies have been done on the IM problem, since his definition by Kempe, Kleinberg, and Tardos (2003). However, most of the work focuses on the non-adaptive version of the problem where all the k seed nodes must be selected before that the cascade starts. In this paper we study the adaptive IM, where the nodes are selected sequentially one by one, and the decision on the i-th seed can be based on the observed cascade produced by the first i-1 seeds. We focus on the full-adoption feedback in which we can observe the entire cascade of each previously selected seed and on the independent cascade model where each edge is associated with an independent probability of diffusing influence. Previous works showed that there are constant upper bounds on the adaptivity gap, which compares the performance of an adaptive algorithm against a non-adaptive one, but the analyses used to prove these bounds only works for specific graph classes such as in-arborescences, out-arborescences, and one-directional bipartite graphs. Our main result is the first sub-linear upper bound that holds for any graph. Specifically, we show that the adaptivity gap is upper-bounded by ∛n+1, where n is the number of nodes in the graph. Moreover we improve over the known upper bound for in-arborescences from 2e/(e-1)≈3.16 to 2e²/(e²-1)≈2.31. Finally, we study α-bounded graphs, a class of undirected graphs in which the sum of node degrees higher than two is at most α, and show that the adaptivity gap is upper-bounded by √α+O(1). Moreover, we show that in 0-bounded graphs, i.e. undirected graphs in which each connected component is a path or a cycle, the adaptivity gap is at most 3e³/(e³-1)≈3.16. To prove our bounds, we introduce new techniques to relate adaptive policies with non-adaptive ones that might be of their own interest.

In the adaptive influence maximization problem, we are given a social network and a budget k, and we iteratively select k nodes, called seeds, in order to maximize the expected number of nodes that are reached by an influence cascade that they generate according to a stochastic model for influence diffusion. The decision on the next seed to select is based on the observed cascade of previously selected seeds. We focus on the myopic feedback model, in which we can only observe which neighbors of previously selected seeds have been influenced and on the independent cascade model, where each edge is associated with an independent probability of diffusing influence. While adaptive policies are strictly stronger than non-adaptive ones, in which all the seeds are selected beforehand, the latter are much easier to design and implement and they provide good approximation factors if the adaptivity gap, the ratio between the adaptive and the non-adaptive optima, is small. Previous works showed that the adaptivity gap is at most 4, and that simple adaptive or non-adaptive greedy algorithms guarantee an approximation of 1/4 (1-1/e) ≈ 0.158 for the adaptive optimum. This is the best approximation factor known so far for the adaptive influence maximization problem with myopic feedback. In this paper, we directly analyze the approximation factor of the non-adaptive greedy algorithm, without passing through the adaptivity gap, and show an improved bound of 1/2 (1-1/e) ≈ 0.316. Therefore, the adaptivity gap is at most 2e/e-1 ≈ 3.164. To prove these bounds, we introduce a new approach to relate the greedy non-adaptive algorithm to the adaptive optimum. The new approach does not rely on multi-linear extensions or random walks on optimal decision trees, which are commonly used techniques in the field. We believe that it is of independent interest and may be used to analyze other adaptive optimization problems. Finally, we also analyze the adaptive greedy algorithm, and show that guarantees an improved approximation factor of 1-1/(√{e)}≈ 0.393.

In polymatrix coordination games, each player x is a node of a graph and must select an action in her strategy set. Nodes are playing separate bimatrix games with their neighbors in the graph. Namely, the utility of x is given by the preference she has for her action plus, for each neighbor y, a payoff which strictly depends on the mutual actions played by x and y. We propose the new class of distance polymatrix coordination games, properly generalizing polymatrix coordination games, in which the overall utility of player x further depends on the payoffs arising by mutual actions of players v,z that are the endpoints of edges at any distance h

We consider a class of coalition formation games that can be succinctly represented by means of hypergraphs and properly generalizes symmetric additively separable hedonic games. More precisely, an instance of hypegraph hedonic game consists of a weighted hypergraph, in which each agent is associated to a distinct node and her utility for being in a given coalition is equal to the sum of the weights of all the hyperedges included in the coalition. We study the performance of stable outcomes in such games, investigating the degradation of their social welfare under two different metrics, the k-Nash price of anarchy and k-core price of anarchy, where k is the maximum size of a deviating coalition. Such prices are defined as the worst-case ratio between the optimal social welfare and the social welfare obtained when the agents reach an outcome satisfying the respective stability criteria. We provide asymptotically tight upper and lower bounds on the values of these metrics for several classes of hypergraph hedonic games, parametrized according to the integer k, the hypergraph arity r and the number of agents n. Furthermore, we show that the problem of computing the exact value of such prices for a given instance is computationally hard, even in case of non-negative hyperedge weights.

We propose and analyse a game describing the interactions between readers and publishers, with the aim of understanding to what extent the strategic behaviour of the latter may influence the quality of content publishing in the World Wide Web. For games with identical publishers, we provide a wide characterization of the cases in which pure Nash equilibria are guaranteed to exist, which mainly depends on the number of publishers and, subordinately, on some of the parameters we use to model their writing abilities. Then, for any game possessing pure Nash equilibria, we show that the price of anarchy is at most 2, even in presence of heterogeneous publishers. Finally, we provide better and tight bounds for some special cases of games with identical publishers.

We build upon previous models for differential pricing in social networks and fair price discrimination in markets, considering a setting in which multiple units of a single product must be sold to selected buyers so as to maximize the seller's revenue or the social welfare, while limiting the differences of the prices offered to social neighbors. We first consider the case of general social graph topologies, and provide optimal or nearly-optimal hardness and approximation results for the related optimization problems under various meaningful assumptions, including the inapproximability within any constant factor on the achievable revenue under the unique game conjecture. Then, we focus on topologies that are typical of social networks. Namely, we consider graphs where the node degrees follow a power-law distribution, and show that it is possible to obtain constant or good approximations for the seller's revenue maximization with high probability, thus improving upon the general case.

We study the existence of allocations of indivisible goods that are envy-free up to one good (EF1), under the additional constraint that each bundle needs to be connected in an underlying item graph G. When the items are arranged in a path, we show that EF1 allocations are guaranteed to exist for arbitrary monotonic utility functions over bundles, provided that either there are at most four agents, or there are any number of agents but they all have identical utility functions. Our existence proofs are based on classical arguments from the divisible cake-cutting setting, and involve discrete analogues of cut-and-choose, of Stromquist's moving-knife protocol, and of the Su-Simmons argument based on Sperner's lemma. Sperner's lemma can also be used to show that on a path, an EF2 allocation exists for any number of agents. Except for the results using Sperner's lemma, all of our procedures can be implemented by efficient algorithms. Our positive results for paths imply the existence of connected EF1 or EF2 allocations whenever G is traceable, i.e., contains a Hamiltonian path. For the case of two agents, we completely characterize the class of graphs G that guarantee the existence of EF1 allocations as the class of graphs whose biconnected components are arranged in a path. This class is strictly larger than the class of traceable graphs; one can check in linear time whether a graph belongs to this class, and if so return an EF1 allocation.

Motivated by possible applications in fault-tolerant routing, we introduce the notion of uniform mixed equilibria in network congestion games with adversarial link failures, where players need to route traffic from a source to a destination node. Given an integer rho >= 1, a rho-uniform mixed strategy is a mixed strategy in which a player plays exactly rho edge disjoint paths with uniform probabilities, so that a rho-uniform mixed equilibrium is a tuple of rho-uniform mixed strategies, one for each player, in which no player can lower her cost by deviating to another rho-uniform mixed strategy. For games with weighted players and affine latency functions, we show existence of rho-uniform mixed equilibria and provide a tight characterization of their price of anarchy. For games with unweighted players, instead, we extend the existential guarantee to any class of latency functions and, restricted to games with affine latencies, we derive a tight characterization of both the prices of anarchy and stability.

To what extent does the structure of the players' strategy space influence the efficiency of decentralized solutions in congestion games? In this work, we investigate whether better performance are possible when restricting to load balancing games in which players can only choose among single resources. We consider three different solutions concepts, namely, approximate pure Nash equilibria, approximate one-round walks generated by selfish players aiming at minimizing their personal cost and approximate one-round walks generated by cooperative players aiming at minimizing the marginal increase in the sum of the players' personal costs. The last two concepts can be interpreted as solutions of greedy online algorithms for the related resource selection problem. We show that, under fairly general latency functions on the resources, better bounds cannot be achieved if players are either weighted or asymmetric. On the positive side, we prove that, under mild assumptions on the latency functions, improvements on the performance of approximate pure Nash equilibria are possible for load balancing games with weighted and symmetric players in the case of identical resources. We also design lower bounds on the performance of one-round walks in load balancing games with unweighted players and identical resources.

We consider the efficiency of taxation in congestion games with polynomial latency functions along the line of research initiated by [Caragiannis et al., ACM Transactions on Algorithms, 2010] who focused on both pure and mixed Nash equilibria in games with affine latencies only. By exploiting the primal-dual method [Bilo, Proceedings of the 10th Workshop on Approximation and Online Algorithms, 2012], we obtain interesting upper bounds with respect to a variety of different solution concepts ranging from approximate pure Nash equilibria up to approximate coarse correlated equilibria, and including also approximate one-round walks starting from the empty state. Our findings show a high beneficial effect of taxation which increases more than linearly with the degree of the latency functions. In some cases, a tight relationship with some well-studied polynomials in Combinatorics and Number Theory, such as the Touchard and the Geometric polynomials, arises. In these cases we can also show matching lower bounds, albeit under mild assumptions; interestingly, our upper bounds are derived by exploiting the combinatorial definition of these polynomials, while our lower bounds are constructed by relying on their analytical characterization.

We investigate the efficiency of some Stackelberg strategies in congestion games with affine latency functions. A Stackelberg strategy is an algorithm that chooses a subset of players and assigns them a prescribed strategy with the purpose of mitigating the detrimental effect that the selfish behavior of the remaining uncoordinated players may cause to the overall performance of the system. The efficiency of a Stackelberg strategy is measured in terms of the price of anarchy of the pure Nash equilibria they induce. Three Stackelberg strategies, namely Largest Latency First , Cover and Scale , were already considered in the literature and non-tight upper and lower bounds on their price of anarchy were given. We reconsider these strategies and provide the exact bound on the price of anarchy of both Largest Latency First and Cover and a better upper bound on the price of anarchy of Scale .