Local strategy learning in networked multi-agent team formation

Bulka, Blazej; Gaston, Matthew E.; desJardins, Marie

doi:10.1007/s10458-006-0007-x

Cited by 28 publications

(15 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The same authors present in (Bulka et al, 2007) an evaluation model for improving the team joining process. In this work, authors embedded agents in fixed network structures and focus on learning team joining policies.…”

Section: Discussionmentioning

confidence: 99%

Challenges for adaptation in agent societies

2012

View full text Add to dashboard Cite

Abstract. Adaptation in Multiagent Systems societies provides a paradign for allowing these societies to change dynamically in order to satisfy the current requirements of the system. This support is especially required for the next generation of systems that focus on open, dynamic, and adaptive applications. In this paper, we analyze the current state of the art regarding approaches that tackle the adaptation issue in these agent societies. We survey the most relevant works up to now in order to highlight the most remarkable feautures according to what they support and how this support is provided. In order to compare these approaches, we also identify different characteristics of the adaptation process that are grouped in different phases. Finally, we discuss some of the most important considerations about the analyzed approaches and we provide some interesting guidelines as open issues that should be required in future developments.

show abstract

Section: Discussionmentioning

confidence: 99%

Challenges for adaptation in agent societies

2012

View full text Add to dashboard Cite

show abstract

“…Similarly, by using social networks of agents, where agents have different skills, and edge weights represent communication costs, the optimal team to complete the task has to cover all the required skills [22,23], or trade off between skills and connectivity [11]. The edges in a social network graph can also be used as constraints, where an agent is assigned a task, and must find teammates that are directly connected to it [7], or form a connected sub-network [5].…”

Section: Team Formationmentioning

confidence: 99%

Weighted synergy graphs for effective team formation with heterogeneous ad hoc agents

Liemhetcharat

Veloso

2014

Artificial Intelligence

View full text Add to dashboard Cite

Previous approaches to select agents to form a team rely on single-agent capabilities, and team performance is treated as a sum of such known capabilities. Motivated by complex team formation situations, we address the problem where both single-agent capabilities may not be known upfront, e.g., as in ad hoc teams, and where team performance goes beyond single-agent capabilities and depends on the specific synergy among agents. We formally introduce a novel weighted synergy graph model to capture new interactions among agents. Agents are represented as vertices in the graph, and their capabilities are represented as Normally-distributed variables. The edges of the weighted graph represent how well the agents work together, i.e., their synergy in a team. We contribute a learning algorithm that learns the weighted synergy graph using observations of performance of teams of only two and three agents. Further, we contribute two team formation algorithms, one that finds the optimal team in exponential time, and one that approximates the optimal team in polynomial time. We extensively evaluate our learning algorithm, and demonstrate the expressiveness of the weighted synergy graph in a variety of problems. We show our approach in a rich ad hoc team formation problem capturing a rescue domain, namely the RoboCup Rescue domain, where simulated robots rescue civilians and put out fires in a simulated urban disaster. We show that the weighted synergy graph outperforms a competing algorithm, thus illustrating the efficacy of our model and algorithms.

show abstract

“…Bulka et al [6] devised a method allowing agents to learn team formation policies for individual agents using a Q-Learning and classifier approach. They showed notable improvement in the performance of their system.…”

Section: Related Workmentioning

confidence: 99%