Fast Solving of Influence Diagrams for Multiagent Planning on GPU-enabled Architectures

Fadel Adoe, Yingke Chen, Prashant Doshi

Abstract

Planning under uncertainty in multiagent settings is highly intractable because of history and plan space complexities. Probabilistic graphical models exploit the structure of the problem domain to mitigate the computational burden. In this paper, we introduce the first parallelization of planning in multiagent settings on a CPU-GPU heterogeneous system. In particular, we focus on the algorithm for exactly solving interactive dynamic influence diagrams, which is a recognized graphical models for multiagent planning. Beyond parallelizing the standard Bayesian inference, the computation of decisions' expected utilities are parallelized. The GPU-based approach provides significant speedup on two benchmark problems.

References

  1. Bernstein, D. S., Givan, R., Immerman, N., and Zilberstein, S. (2002). The complexity of decentralized control of Markov decision processes. Mathematics of Operations Research, 27(4):819-840.
  2. Berstein, D. S., Hansen, E. A., and Zilberstein, S. (2005). Bounded policy iteration for decentralized POMDPs. In IJCAI, pages 1287-1292.
  3. Chandrasekaran, M., Doshi, P., Zeng, Y., and Chen, Y. (2014). Team behavior in interactive dynamic influence diagrams with applications to ad hoc teams. In AAMAS, pages 1559-1560.
  4. Chen, Y., Hong, J., Liu, W., Godo, L., Sierra, C., and Loughlin, M. (2013). Incorporating PGMs into a BDI architecture. In PRIMA, pages 54-69.
  5. Doshi, P., Zeng, Y., and Chen, Q. (2009). Graphical models for interactive POMDPs: Representations and solutions. JAAMAS, 18(3):376-416.
  6. Gal, K. and Pfeffer, A. (2008). Networks of influence diagrams: A formalism for representing agents' beliefs and decision-making processes. JAIR, 33:109-147.
  7. Gmytrasiewicz, P. J. and Doshi, P. (2005). A framework for sequential planning in multiagent settings. JAIR, 24:49-79.
  8. Howard, R. A. and Matheson, J. E. (1984). Influence diagrams. In Howard, R. A. and Matheson, J. E., editors, The Principles and Applications of Decision Analysis. Strategic Decisions Group, Menlo Park, CA 94025.
  9. Jeon, H., Xia, Y., and Prasanna, K. V. (2010). Parallel exact inference on a cpu-gpgpu heterogenous system. In ICPP, pages 61-70.
  10. Koller, D. and Friedman, N. (2009). Probabilistic Graphical Models: Principles and Techniques. MIT Press.
  11. Koller, D. and Milch, B. (2001). Multi-agent influence diagrams for representing and solving games. In IJCAI, pages 1027-1034.
  12. Luo, J., Yin, H., Li, B., and Wu, C. (2011). Path planning for automated guided vehicles system via I-DIDs with communication. In ICCA, pages 755 -759.
  13. Pearl, J. (1998). Probabilistic Reasoning in Intelligent Systems: Networks of Plausible Inference. Morgan Kaufmann, Berlin, Germany.
  14. Sano, Y., Kadono, Y., and Fukuta, N. (2014). A performance optimization support framework for gpu-based traffic simulations with negotiating agents. In ACAN.
  15. Smallwood, R. and Sondik, E. (1973). The optimal control of partially observable Markov decision processes over a finite horizon. Operations Research, 21:1071- 1088.
  16. S√łndberg-Jeppesen, N., Jensen, F. V., and Zeng, Y. (2013). Opponent modeling in a PGM framework. In AAMAS, pages 1149-1150.
  17. V. Kozlov, A. and Pal Singh, J. (1994). A parallel LauritzenSpiegelhalter algorithm for probabilistic inference. In Supercomputing, pages 320-329.
  18. Xia, Y. and Prasanna, K. V. (2008). Parallel exact inference on the cell broadband engine processor. In SC, pages 1-12.
  19. Zeng, Y. and Doshi, P. (2012). Exploiting model equivalences for solving interactive dynamic influence diagrams. JAIR, 43:211-255.
  20. Zheng, L., Mengshoel, O. J., and Chong, J. (2011). Belief propagation by message passing in junction trees: Computing each message faster using gpu parallelization. In UAI.
Download


Paper Citation


in Harvard Style

Adoe F., Chen Y. and Doshi P. (2015). Fast Solving of Influence Diagrams for Multiagent Planning on GPU-enabled Architectures . In Proceedings of the International Conference on Agents and Artificial Intelligence - Volume 2: ICAART, ISBN 978-989-758-074-1, pages 183-195. DOI: 10.5220/0005224001830195


in Bibtex Style

@conference{icaart15,
author={Fadel Adoe and Yingke Chen and Prashant Doshi},
title={Fast Solving of Influence Diagrams for Multiagent Planning on GPU-enabled Architectures},
booktitle={Proceedings of the International Conference on Agents and Artificial Intelligence - Volume 2: ICAART,},
year={2015},
pages={183-195},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0005224001830195},
isbn={978-989-758-074-1},
}


in EndNote Style

TY - CONF
JO - Proceedings of the International Conference on Agents and Artificial Intelligence - Volume 2: ICAART,
TI - Fast Solving of Influence Diagrams for Multiagent Planning on GPU-enabled Architectures
SN - 978-989-758-074-1
AU - Adoe F.
AU - Chen Y.
AU - Doshi P.
PY - 2015
SP - 183
EP - 195
DO - 10.5220/0005224001830195