Name
Title | ||
|---|---|---|
Letting loose a SPIDER on a network of POMDPs: generating quality guaranteed policies |
Abstract | ||
|---|---|---|
Distributed Partially Observable Markov Decision Problems (Distributed POMDPs) are a popular approach for modeling multi-agent systems acting in uncertain domains. Given the significant complexity of solving distributed POMDPs, particularly as we scale up the numbers of agents, one popular approach has focused on approximate solutions. Though this approach is efficient, the algorithms within this approach do not provide any guarantees on solution quality. A second less popular approach focuses on global optimality, but typical results are available only for two agents, and also at considerable computational cost. This paper overcomes the limitations of both these approaches by providing SPIDER, a novel combination of three key features for policy generation in distributed POMDPs: (i) it exploits agent interaction structure given a network of agents (i.e. allowing easier scale-up to larger number of agents); (ii) it uses a combination of heuristics to speedup policy search; and (iii) it allows quality guaranteed approximations, allowing a systematic tradeoff of solution quality for time. Experimental results show orders of magnitude improvement in performance when compared with previous global optimal algorithms. |
Year | DOI | Venue |
|---|---|---|
2007 | 10.1145/1329125.1329388 | AAMAS |
Keywords | Field | DocType |
previous global optimal algorithm,approximate solution,popular approach,global optimality,policy generation,novel combination,speedup policy search,solution quality,partially observable markov decision,agent interaction structure,multi agent system,global optimization,multi agent systems | Computer science,Multi-agent system,Heuristics,Artificial intelligence,Speedup,Distributed computing,Decision problem,Mathematical optimization,Partially observable Markov decision process,Markov chain,Exploit,Global optimality,Machine learning | Conference |
Citations | PageRank | References |
28 | 1.08 | 11 |
Authors | ||
5 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Pradeep Varakantham | 1 | 648 | 63.05 |
| Janusz Marecki | 2 | 685 | 49.06 |
| Yuichi Yabu | 3 | 28 | 2.09 |
| Milind Tambe | 4 | 6008 | 522.25 |
| Makoto Yokoo | 5 | 3632 | 421.99 |