Scheduler
eligibilityTraceAlgo.h
Go to the documentation of this file.
1 
10 #ifndef ELIGIBILITY_TRACE_ALGO_H
11 #define ELIGIBILITY_TRACE_ALGO_H
12 
13 #include "rlBackupAlgorithm.h"
15 
16 namespace Mdp
17 {
18 
20 {
21 public:
22  EligibilityTraceAlgo(std::shared_ptr<Context> c, TabularActionValues *av)
23  : RlBackupAlgorithm(c, av)
24  , tabularAv(av)
25  {};
26  virtual void init();
27  virtual void end();
28  virtual void updateActionValues(state_t previousState, state_t nextState, action_t previousAction, double reward)=0;
29 protected:
30  void updateState(state_t previousState, action_t previousAction, double reward);
34  double previousReward{0.0};
35  std::vector<std::vector<double>> e;
36  double lambda{0.5};
37  double discountFactor{0.5};
38  size_t stateSize{0};
39  size_t actionSize{0};
40 };
41 
42 
43 }
44 
45 #endif
TabularActionValues * tabularAv
void updateState(state_t previousState, action_t previousAction, double reward)
std::vector< std::vector< double > > e
virtual void updateActionValues(state_t previousState, state_t nextState, action_t previousAction, double reward)=0
size_t action_t
Definition: action_impl.h:18
Definition: action.h:18
EligibilityTraceAlgo(std::shared_ptr< Context > c, TabularActionValues *av)
size_t state_t
Definition: state.h:19
Definition: reward.py:1