Since the pioneering work by R. Coulom (2006), Monte-Carlo Tree Search is an important tool for discrete time control. Nonetheless, there are clear weaknesses to be corrected by further work.



Already done:
  • Generalities
    • random processes (A. Takahashi)
    • introduction into the Mash git repository ( )
    • Interfacing the Mash problems with our solvers (J.-B. Hoock, October 2010)
  • Solvers
    • preliminary MCTS solver (A. Couetoux, O. Teytaud). Update: now stable.
    • various simple solvers for comparison (J.-B. Hoock, N. Sokolovska, O. Teytaud).
    • a decision stump solver (learning by direct policy search; J.-B. Hoock).
    • preliminary Q-learning implementation (N. Sokolovska)
    • OpenDP solvers are under integration (O. Teytaud; first integration November 2010). Update: cancelled, too many problems.
    • Direct Policy Search, many families of functions.
  • Testbeds
    • Swing problem (A. Takahashi, O. Teytaud)
    • Stock management problem (Artelys + A. Couetoux + O. Teytaud)
    • Energy management problem (but: not yet interfaced with the rest of the platform). Update (March 2013): interfacing ok.
    • OpenDP problems are under integration (O. Teytaud). Update: replaced by Mash platform.


  • UCT in continuous domains
  • January 2010: paper on MCTS for power systems accepted in Roadef 2011.
  • February 2010: the Roadef paper on MCTS for stock management is preselected for student award.
  • publications by A. Couetoux on Monte-Carlo Tree Search and Direct Policy Search for energy management.

More informations ? email me

Photos ? see here.