RL.TD

Synopsis

• data Q_Opts = Q_Opts {
  • o_alpha :: TD_Number
  • o_gamma :: TD_Number
  • o_eps :: TD_Number
  }
• defaultOpts :: Q_Opts
• type TD_Number = Double
• type Q s a = M s a TD_Number
• emptyQ :: TD_Number -> Q s a
• toV :: (Bounded a, Enum a, Eq a, Hashable a, Eq s, Hashable s) => Q s a -> HashMap s TD_Number
• class (Monad m, Eq s, Hashable s, Show s, Eq a, Hashable a, Enum a, Bounded a, Show a) => TD_Problem pr m s a | pr -> m, pr -> s, pr -> a where
• queryQ :: (Hashable s, Hashable k, MonadState (M s k v) f, Eq s, Eq k, Enum k, Bounded k) => s -> f [(k, v)]
• modifyQ :: (TD_Problem pr m s a, MonadTrans t, MonadState (Q s a) (t m)) => pr -> s -> a -> (TD_Number -> TD_Number) -> t m ()
• action :: (TD_Problem pr m1 s1 a, MonadRnd g m, Hashable s, MonadState (M s a num) m, Real num, Fractional num, Eq s) => pr -> s -> num -> m (a, num)
• transition :: (TD_Problem pr m b a, MonadTrans t, MonadState (Q b a) (t m)) => pr -> b -> a -> t m b
• q_learn :: (MonadRnd g m, TD_Problem pr m s a) => Q_Opts -> Q s a -> s -> pr -> m (s, Q s a)
• q_exec :: (MonadRnd g m, TD_Problem pr m s a) => Q_Opts -> Q s a -> s -> pr -> m s

Documentation