Georgios Chalkiadakis

Bayesian Reinforcement Learning for Coalition Formation under Uncertainty

Research on coalition formation usually assumes the values of potential coalitions to be known with certainty. Furthermore, settings in which agents lack sufficient knowledge of the capabilities of potential partners is rarely, if ever, touched upon. We remove these often unrealistic assumptions and propose a model that utilizes Bayesian (multiagent) reinforcement learning in a way that enables coalition participants to reduce their uncertainty regarding coalitional values and the capabilities of others. In addition, we introduce the Bayesian Core, a new stability concept for coalition formation under uncertainty. Preliminary experimental evidence demonstrates the effectiveness of our approach.

Cooperative games with overlapping coalitions

In the usual models of cooperative game theory, the outcome of a coalition formation process is either the grand coalition or a coalition structure that consists of disjoint coalitions. However, in many domains where coalitions are associated with tasks, an agent may be involved in executing more than one task, and thus may distribute his resources among several coalitions. To tackle such scenarios, we introduce a model for cooperative games with overlapping coalitions-or overlapping coalition formation (OCF) games. We then explore the issue of stability in this setting. In particular, we introduce a notion of the core, which generalizes the corresponding notion in the traditional (non-overlapping) scenario. Then, under some quite general conditions, we characterize the elements of the core, and show that any element of the core maximizes the social welfare. We also introduce a concept of balancedness for overlapping coalitional games, and use it to characterize coalition structures that can be extended to elements of the core. Finally, we generalize the notion of convexity to our setting, and show that under some natural assumptions convex games have a non-empty core. Moreover, we introduce two alternative notions of stability in OCF that allow a wider range of deviations, and explore the relationships among the corresponding definitions of the core, as well as the classic (non-overlapping) core and the Aubin core. We illustrate the general properties of the three cores, and also study them from a computational perspective, thus obtaining additional insights into their fundamental structure.

LigaSure versus clamp‐and‐tie thyroidectomy for benign nodular disease

LigaSure is an alternative bipolar diathermy system (BDS) combining vessel sealing with reduced thermal spread, which has been successfully used in many abdominal operations; however, there is a little experience in open thyroidectomy. This study compares the efficacy and the advantages of the LigaSure BDS when used for total thyroidectomy for benign thyroid disease with the conventional clamp‐and‐tie technique.

Coalition formation under uncertainty: bargaining equilibria and the Bayesian core stability concept

Coalition formation is a problem of great interest in AI, allowing groups of autonomous, rational agents to form stable teams. Furthermore, the study of coalitional stability concepts and their relation to equilibria that guide the strategic interactions of agents during bargaining has lately attracted much attention. However, research to date in both AI and economics has largely ignored the potential presence of uncertainty when studying either coalitional stability or coalitional bargaining. This paper is the first to relate a (cooperative) stability concept under uncertainty, the Bayesian core (BC), with (non-cooperative) equilibrium concepts of coalitional bargaining games. We prove that if the BC of a coalitional game (and of each subgame) is non-empty, then there exists an equilibrium of the corresponding bargaining game that produces a BC element; and conversely, if there exists a coalitional bargaining equilibrium (with certain properties), then it induces a BC configuration. We thus provide a non-cooperative justification of the BC stability concept. As a corollary, we establish a sufficient condition for the existence of the BC. Finally, for small games, we provide an algorithm to decide whether the BC is non-empty.

Sequentially optimal repeated coalition formation under uncertainty

Coalition formation is a central problem in multiagent systems research, but most models assume common knowledge of agent types. In practice, however, agents are often unsure of the types or capabilities of their potential partners, but gain information about these capabilities through repeated interaction. In this paper, we propose a novel Bayesian, model-based reinforcement learning framework for this problem, assuming that coalitions are formed (and tasks undertaken) repeatedly. Our model allows agents to refine their beliefs about the types of others as they interact within a coalition. The model also allows agents to make explicit tradeoffs between exploration (forming “new” coalitions to learn more about the types of new potential partners) and exploitation (relying on partners about which more is known), using value of information to define optimal exploration policies. Our framework effectively integrates decision making during repeated coalition formation under type uncertainty with Bayesian reinforcement learning techniques. Specifically, we present several learning algorithms to approximate the optimal Bayesian solution to the repeated coalition formation and type-learning problem, providing tractable means to ensure good sequential performance. We evaluate our algorithms in a variety of settings, showing that one method in particular exhibits consistently good performance in practice. We also demonstrate the ability of our model to facilitate knowledge transfer across different dynamic tasks.

Overlapping Coalition Formation

In multiagent domains, agents form coalitions to perform tasks. The usual models of cooperative game theory assume that the desired outcome is either the grand coalition or a coalition structure that consists of disjoint coalitions (i.e., a partition of the set of agents). However, in practice an agent may be involved in executing more than one task, and distributing his resources between several (not necessarily disjoint) coalitions. To tackle such scenarios, we introduce a model for cooperative games with overlapping coalitions. We then focus on concepts of stability in this setting. In particular, we define and study a notion of the core, which is a generalization of the corresponding notion in the traditional models of cooperative game theory. Under some quite general conditions, we characterize the elements of core. As a corollary, we also show that any element of the core maximizes the social welfare. We then introduce a concept of balancedness for overlapping coalitional games, and use it to characterize coalition structures that can be extended to elements of the core. Furthermore, we generalize the notion of convexity to our setting, and show that under some natural assumptions convex games have a non-empty core. To the best of our knowledge, this is the first paper to provide a generic model for overlapping coalition formation, along with a theoretical treatment of stability in this setting.

Cooperative Game Theory: Basic Concepts and Computational Challenges

Cooperative game theory studies situations in which agents can benefit by working together. This article outlines the key concepts of cooperative game theory, and discusess the challenges that arise in applying these in AI applications.

Cooperatives for demand side management

We propose a new scheme for efficient demand side management for the Smart Grid. Specifically, we envisage and promote the formation of cooperatives of medium-large consumers and equip them (via our proposed mechanisms) with the capability of regularly participating in the existing electricity markets by providing electricity demand reduction services to the Grid. Based on mechanism design principles, we develop a model for such cooperatives by designing methods for estimating suitable reduction amounts, placing bids in the market and redistributing the obtained revenue amongst the member agents. Our mechanism is such that the member agents have no incentive to show artificial reductions with the aim of increasing their revenues.

Characteristic function games with restricted agent interactions

In many real-world settings, the structure of the environment constrains the formation of coalitions among agents. These settings can be represented by characteristic function games, also known as coalitional games, equipped with interaction graphs. An interaction graph determines the set of all feasible coalitions, in that a coalition C can form only if the subgraph induced over the nodes/agents in C is connected. Our work analyzes stability issues arising in such environments, by focusing on the core as a solution concept, and by considering the coalition structure viewpoint, that is, without assuming that the grand-coalition necessarily forms.The complexity of the coalition structure core is studied over a variety of interaction graph structures of interest, including complete graphs, lines, cycles, trees, and nearly-acyclic graphs (formally, having bounded treewidth). The related stability concepts of the least core and the cost of stability are also studied. Results are derived for the setting of compact coalitional games, i.e., for games that are implicitly described via a compact encoding, and where simple calculations on this encoding are to be performed in order to compute the payoff associated with any coalition. Moreover, specific results are provided for compact games defined via marginal contribution networks, an expressive encoding mechanism that received considerable attention in the last few years.

Predicting the power output of distributed renewable energy resources within a broad geographical region

In recent years, estimating the power output of inherently intermittent and potentially distributed renewable energy sources has become a major scientific and societal concern. In this paper, we provide an algorithmic framework, along with an interactive web-based tool, to enable short-to-middle term forecasts of photovoltaic (PV) systems and wind generators output. Importantly, we propose a generic PV output estimation method, the backbone of which is a solar irradiance approximation model that incorporates free-to-use, readily available meteorological data coming from online weather stations. The model utilizes non-linear approximation components for turning cloud-coverage into radiation forecasts, such as an MLP neural network with one hidden layer. We present a thorough evaluation of the proposed techniques, and show that they can be successfully employed within a broad geographical region (the Mediterranean belt) and come with specific performance guarantees. Crucially, our methods do not rely on complex and expensive weather models and data, and our web-based tool can be of immediate use to the community as a simulation data acquisition platform.