Module QlearnSource

Sourcetype ('state, 'action) t
Sourceval make : ?rate:float -> ?discount:float -> ?explore:float -> ?seed:int -> unit -> ('state, 'action) t
Sourceval value : ('state, 'action) t -> 'state -> 'action -> float
Sourceval best : ('state, 'action) t -> 'state -> 'action list -> 'action option
Sourceval choose : ('state, 'action) t -> 'state -> 'action list -> 'action option
Sourceval learn : ('state, 'action) t -> state:'state -> action:'action -> reward:float -> next:'state -> next_actions:'action list -> unit
Sourcetype ('state, 'action) world = {
  1. actions : 'state -> 'action list;
  2. step : 'state -> 'action -> 'state * float;
  3. over : 'state -> bool;
}
Sourceval episode : ?limit:int -> ('state, 'action) t -> ('state, 'action) world -> 'state -> float
Sourceval greedy_run : ?limit:int -> ('state, 'action) t -> ('state, 'action) world -> 'state -> float * 'state list
Sourceval known : ('state, 'action) t -> int
Sourceval values : ('state, 'action) t -> 'state -> 'action list -> ('action * float) list