Cooperative game theory is a branch of (micro-)economics that studies the behavior self-interested agents in strategic settings where binding agreements among are possible. Our aim this b

Much emphasis in multiagent reinforcement learning (MARL) research is placed on ensuring that MARL algorithms (eventually) converge to desirable equilibria. As standard learning, convergence generally requires sufficient exploration of strategy space. However, often comes at a price the form penalties or foregone opportunities. In settings, problem exacerbated by need for agents "coordinate" their policies We propose Bayesian model optimal problems allows these costs be weighed against...

In the usual models of cooperative game theory, outcome a coalition formation process is either grand or structure that consists disjoint coalitions. However, in many domains where coalitions are associated with tasks, an agent may be involved executing more than one task, and thus distribute his resources among several To tackle such scenarios, we introduce model for games overlapping coalitions--or (OCF) games. We then explore issue stability this setting. particular, notion core, which...

The creation of Virtual Power Plants (VPPs) has been suggested in recent years as the means for achieving cost-efficient integration many distributed energy resources (DERs) that are starting to emerge electricity network. In this work, we contribute development VPPs by offering a game-theoretic perspective problem. Specifically, design cooperatives (or cooperative VPPs---CVPPs) rational autonomous DER-agents representing small-to-medium size renewable producers, which coalesce profitably...

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

In this paper, we develop a novel hybrid recommender system for the tourism domain, which combines (a) Bayesian preferences elicitation component operates by asking user to rate generic images (corresponding types of POIs) in order build model and (b) content-based (CB) recommendations component. The second can fact itself be considered among two different CB algorithms, each exploiting one semantic similarity measures: hierarchy-based non-hierarchy based one. latter is recently introduced...

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

We present the first real-world benchmark for sequentially-optimal team formation, working within framework of a class online football prediction games known as Fantasy Football. model problem Bayesian reinforcement learning one, where action space is exponential in number players and decision maker's beliefs are over multiple characteristics each footballer. then exploit domain knowledge to construct computationally tractable solution techniques order build competitive automated Football...

Lane-free traffic is a novel research domain, in which vehicles no longer adhere to the notion of lanes, and consider whole lateral space within road boundaries. This constitutes an entirely different problem domain for autonomous driving compared lane-based traffic, as there leader vehicle or lane-changing operation. Therefore, observations need properly accommodate lane-free environment without carrying over bias from approaches. The recent successes deep reinforcement learning (DRL)...

In this paper we present a novel approach for multiagent decision making in dynamic environments based on Factor Graphs and the Max-Sum algorithm, considering asynchronous variable reassignments distributed message-passing among agents. Motivated by challenging domain of lane-free traffic where automated vehicles can communicate coordinate as agents, propose more realistic communication framework Graph formulations that satisfies above-mentioned restrictions, along with Conditional Max-Sum:...

The growing focus on sustainable and environmentally friendly energy production has resulted in the proliferation of distributed resources (DERs), mainly based renewable sources like wind sunlight. However, their small size intermittent nature supply means that such generators cannot easily be assimilated into current electricity network (Grid) conventional generators. Against this background, Virtual Power Plants are fast emerging as a solution to problem whereby large number may aggregated...

Coalition formation is a problem of great interest in AI, allowing groups autonomous, rational agents to form stable teams. Furthermore, the study coalitional stability concepts and their relation equilibria that guide strategic interactions during bargaining has lately attracted much attention. However, research date both AI economics largely ignored potential presence uncertainty when studying either or bargaining. This paper first relate (cooperative) concept under uncertainty, Bayesian...

In this paper, we present a directly applicable scheme for electricity consumption shifting and effective demand curve flattening. The can employ the services of either individual or cooperating consumer agents alike. Agents participating in scheme, however, are motivated to form cooperatives, order reduce their bills via lower group prices granted sizable from high low time intervals. takes into account costs, uses strictly proper scoring rule reward contributors according efficiency....

In this paper, we deal with the sequential decision making problem of agents operating in computational economies, where there is uncertainty regarding trustworthiness service providers populating environment. Specifically, propose a generic Bayesian trust model, and formulate optimal solution to exploration-exploitation facing when repeatedly interacting others such environments. We then present computationally tractable reinforcement learning algorithm approximate that by taking into...

Cooperative games with overlapping coalitions (OCF games) [3, 23] model scenarios where agents can distribute their resources among several tasks; each task generates a profit which may be freely divided the participating in task. The goal of this work is to initiate systematic investigation algorithmic aspects OCF games. We propose discretized over-lapping coalition formation, agent i e N has weight wi and allocate an integer amount any Within framework, we focus on computation outcomes...

This paper presents the ongoing development of microscopic TrafficFluid-Sim simulator, aimed primarily for Connected and Automated Vehicles (CAVs) under a novel lane-free traffic paradigm. In particular, builds on SUMO simulation infrastructure to model environments featuring two vehicle characteristics: (i) can be located at any arbitrary lateral position within road boundaries; (ii) may exert, based their automated driving connectivity capabilities, "vehicle nudging" other surrounding...

The problem of coalition formation when agents are uncertain about the types or capabilities their potential partners is a critical one. In [3] Bayesian reinforcement learning framework developed for this coalitions formed (and tasks undertaken) repeatedly: not only does model allow to refine beliefs others, but uses value information define optimal exploration policies. However, computational approximations in that work purely myopic. We present novel, non-myopic algorithms approximate...