Results 1 
7 of
7
Asynchronous Algorithms for Approximate Distributed Constraint Optimization with Quality Bounds
"... Distributed Constraint Optimization (DCOP) is a popular framework for cooperative multiagent decision making. DCOP is NPhard, so an important line of work focuses on developing fast incomplete solution algorithms for largescale applications. One of the few incomplete algorithms to provide bounds o ..."
Abstract

Cited by 23 (5 self)
 Add to MetaCart
Distributed Constraint Optimization (DCOP) is a popular framework for cooperative multiagent decision making. DCOP is NPhard, so an important line of work focuses on developing fast incomplete solution algorithms for largescale applications. One of the few incomplete algorithms to provide bounds on solution quality is ksize optimality, which defines a local optimality criterion based on the size of the group of deviating agents. Unfortunately, the lack of a generalpurpose algorithm and the commitment to forming groups based solely on group size has limited the use of ksize optimality. This paper introduces tdistance optimality which departs from ksize optimality by using graph distance as an alternative criteria for selecting groups of deviating agents. This throws open a new research direction into the tradeoffs between different group selection
Local Optimal Solutions for DCOP: New Criteria, Bound, and Algorithm
"... Distributed constraint optimization (DCOP) is a popular formalism for modeling cooperative multiagent systems. In largescale networks, finding a global optimum using complete algorithms is often impractical, which leads to the study on incomplete algorithms. Traditionally incomplete algorithms c ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
(Show Context)
Distributed constraint optimization (DCOP) is a popular formalism for modeling cooperative multiagent systems. In largescale networks, finding a global optimum using complete algorithms is often impractical, which leads to the study on incomplete algorithms. Traditionally incomplete algorithms can only find locally optimal solution with no quality guarantees. Recent work on ksizeoptimality has established bounds on solution quality, but size is not the only criteria for forming local optimization groups. In addition, there is only one algorithm for computing solutions for arbitrary k and it is quite inefficient. We introduce tdistanceoptimality, which offers an alternative way to specify optimization groups. We establish bounds for this criteria that are often tighter than those for koptimality. We then introduce an asynchronous local search algorithm for tdistanceoptimality. We implement and evaluate the algorithm for both t and k optimality that offer significant improvements over KOPT – the only existing algorithm for ksizeoptimality. Our experiment shows tdistanceoptimality converges more quickly and to better solutions than ksizeoptimality in scalefree graphs, but ksizeoptimality has advantages for random graphs. 1.
Two decades of multiagent teamwork research: Past, present, and future
, 2010
"... Abstract. This paper discusses some of the recent cooperative multiagent systems ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
(Show Context)
Abstract. This paper discusses some of the recent cooperative multiagent systems
A Scalable Algorithm to Solve Distributed Constraint Optimization
"... Abstract Recently, Distributed Constraint Optimization Problems (DCOP) have been drawing a growing body of attention as an important research area in multi agent systems as a large body of real problems can be modeled by them. The primary goal of this research is to design a distributed and effecti ..."
Abstract
 Add to MetaCart
(Show Context)
Abstract Recently, Distributed Constraint Optimization Problems (DCOP) have been drawing a growing body of attention as an important research area in multi agent systems as a large body of real problems can be modeled by them. The primary goal of this research is to design a distributed and effective algorithm to solve DCOP. There are various criteria that measure the efficiency of DCOP algorithms, but the most efficient algorithm for DCOP is the one by which the computation and communication cost is as low as possible and the quality of the solution is high. In this paper, we focus on an approximate DCOP algorithm called DALO (Distributed Asynchronous Local Optimization). Using the main idea of the DALO algorithm, we propose a new algorithm to solve DCOP, which exhibits two important improvements over the DALO algorithm. First we use a sequential partial approach to select a coefficient of leaders to compute the best assignment for agents by which the computation and communication cost decrease in the whole DCOP. The second improvement is an evolutionary approach by which the computation and communication burden for each agent decreases. We present some empirical evidences that show our algorithm performs better than the DALO algorithm. key words distributed constraint optimization, multi agent system I
Pseudotreebased Incomplete Algorithm for Distributed Constraint Optimization with Quality Bounds
"... Abstract. A Distributed Constraint Optimization Problem (DCOP) is a fundamental problem that can formalize various applications related to multiagent cooperation. Since it is NPhard, considering faster incomplete algorithms is necessary for largescale applications. Most incomplete algorithms gen ..."
Abstract
 Add to MetaCart
(Show Context)
Abstract. A Distributed Constraint Optimization Problem (DCOP) is a fundamental problem that can formalize various applications related to multiagent cooperation. Since it is NPhard, considering faster incomplete algorithms is necessary for largescale applications. Most incomplete algorithms generally do not provide any guarantees on the quality of solutions. Some notable exceptions are DALO, the bounded maxsum algorithm, and ADPOP. In this paper, we develop a new solution criterion called poptimality and an incomplete algorithm for obtaining a poptimal solution. The characteristics of this algorithm are as follows: (i) it can provide the upper bounds of the absolute/relative errors of the solution, which can be obtained a priori/a posteriori, respectively, (ii) it is based on a pseudotree, which is a widely used graph structure in complete DCOP algorithms, (iii) it is a oneshot type algorithm, which runs in polynomialtime in the number of agents n, and (iv) it has adjustable parameter p, so that agents can tradeoff better solution quality against computational overhead. The evaluation results illustrate that this algorithm can obtain better quality solutions and bounds compared to existing bounded incomplete algorithms, while the run time of this algorithm is shorter.
Towards a Theoretic Understanding of DCEE
"... Abstract. Common wisdom says that the greater the level of teamwork, the higher the performance of the team. In teams of cooperative autonomous agents, working together rather than independently can increase the team reward. However, recent results show that in uncertain environments, increasing the ..."
Abstract
 Add to MetaCart
(Show Context)
Abstract. Common wisdom says that the greater the level of teamwork, the higher the performance of the team. In teams of cooperative autonomous agents, working together rather than independently can increase the team reward. However, recent results show that in uncertain environments, increasing the level of teamwork can actually decrease overall performance. Coined the team uncertainty penalty, this phenomenon has been shown empirically in simulation, but the underlying mathematics are not yet understood. By understanding the mathematics, we could develop algorithms that reduce or eliminate this penalty of increased teamwork. In this paper we investigate the team uncertainty penalty on two fronts. First, we provide results of robots exhibiting the same behavior seen in simulations. Second, we present a mathematical foundation by which to analyze the phenomenon. Using this model, we present findings indicating that the team uncertainty penalty is inherent to the level of teamwork allowed, rather than to specific algorithms. 1
Algorithms, Experimentation
"... Increasing teamwork between agents typically increases the performance of a multiagent system, at the cost of increased communication and higher computational complexity. This work examines joint actions in the context of a multiagent optimization problem where agents must cooperate to balance exp ..."
Abstract
 Add to MetaCart
(Show Context)
Increasing teamwork between agents typically increases the performance of a multiagent system, at the cost of increased communication and higher computational complexity. This work examines joint actions in the context of a multiagent optimization problem where agents must cooperate to balance exploration and exploitation. Surprisingly, results show that increased teamwork can hurt agent performance, even when communication and computation costs are ignored, which we term the team uncertainty penalty. This paper introduces the above phenomena, analyzes it, and presents algorithms to reduce the effect of the penalty in our problem setting.