ast_toolbox.algos.mctsbv module

class ast_toolbox.algos.mctsbv.MCTSBV(M=10, **kwargs)[source]

Bases: ast_toolbox.algos.mcts.MCTS

Monte Carlo Tress Search (MCTS) with double progressive widening (DPW) [1] using Blind Value search from Couetoux et al. [2].

Parameters:
  • M (int, optional) – The number of randon decisions generated for the action pool.
  • kwargs – Keyword arguments passed to ast_toolbox.algos.mcts.MCTS.

References

[1]Lee, Ritchie, et al. “Adaptive stress testing of airborne collision avoidance systems.” 2015 IEEE/AIAA 34th Digital Avionics Systems Conference (DASC). IEEE, 2015.
[2]Couetoux, Adrien, Hassen Doghmen, and Olivier Teytaud. “Improving the exploration in upper confidence trees.” International Conference on Learning and Intelligent Optimization. Springer, Berlin, Heidelberg, 2012.
init()[source]

Initiate AST internal parameters