A Recursive Classifier System for Partially Observable Environments

@article{Hamzeh2009ARC,
  title={A Recursive Classifier System for Partially Observable Environments},
  author={Ali Hamzeh and Sattar Hashemi and Ashkan Sami and Adel Torkaman Rahmani},
  journal={Fundam. Inform.},
  year={2009},
  volume={97},
  pages={15-40}
}
Previously we introduced Parallel Specialized XCS (PSXCS) , a distributed-architecture classifier system that detects aliased environmental state s and assigns their handling to created subordinate XCS classifier systems. PSXCS uses a history-windo approach, but with novel efficiency since the subordinate XCSs, which employ the windows, are on ly spawned for parts of the state space that are actually aliased. However, because the window leng ths are finite and set manually, PSXCS may fail to be… CONTINUE READING

Similar Papers

Loading similar papers…