|
||||||||||
PREV CLASS NEXT CLASS | FRAMES NO FRAMES | |||||||||
SUMMARY: NESTED | FIELD | CONSTR | METHOD | DETAIL: FIELD | CONSTR | METHOD |
java.lang.Objectjmarkov.jmdp.MDP<S,A>
jmarkov.jmdp.InfiniteMDP<S,A>
S
- States class.A
- Actions class.public abstract class InfiniteMDP<S extends State,A extends Action>
This class is a structural class and is. It represents a general Infinite horizon MDP problem. It is extended for discrte and continuous problems.
PolicyIterationSolver
,
ValueIterationSolver
Field Summary | |
---|---|
protected States<S> |
absorbingStates
Set of absorving states. |
protected long |
explorationTime
Time used to explore the system. |
protected boolean |
hasAbsorbingState
Whether an absorving state was found |
protected int |
numStates
Number of states. |
protected ValueFunction<S> |
probability
The value function |
protected ProbabilitySolver<S,A> |
probabilitySolver
Idiotic solver that was of course not needed since JMarkov can handle the job |
protected StatesSet<S> |
states
set of states |
Fields inherited from class jmarkov.jmdp.MDP |
---|
finite, initial, reporter |
Constructor Summary | |
---|---|
InfiniteMDP(States<S> initial)
Creates a new INFINITE Dynamic Programming (DP) Problem. |
Method Summary | |
---|---|
abstract Actions<A> |
feasibleActions(S i)
Returns the set of actions available at this state. |
protected abstract StatesSet<S> |
generate()
|
StatesSet<S> |
getAllStates()
Complete set of states explored |
protected AbstractAverageSolver<S,A> |
getDefaultAverageSolver()
|
protected AbstractDiscountedSolver<S,A> |
getDefaultDiscountedSolver(double interestRate)
|
protected AbstractInfiniteSolver<S,A> |
getDefaultSolver()
The class that extends MDP must define the default solver to use. |
int |
getNumStates()
Returns the number of states in the model. |
AbstractInfiniteSolver<S,A> |
getSolver()
|
protected void |
setInterestRate(double interestRate)
Sets the interest rate to be used in the problem solving if the objective is to minimze the discounted cost. |
Methods inherited from class jmarkov.jmdp.MDP |
---|
debug, debug, debug, getDebugLevel, getOptimalPolicy, getOptimalValueFunction, getReporter, isFinite, isSolved, operation, printSolution, printSolution, setDebugLevel, setReporter, setSolver, solve |
Methods inherited from class java.lang.Object |
---|
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait |
Field Detail |
---|
protected StatesSet<S extends State> states
protected ValueFunction<S extends State> probability
protected ProbabilitySolver<S extends State,A extends Action> probabilitySolver
protected boolean hasAbsorbingState
protected States<S extends State> absorbingStates
protected long explorationTime
protected int numStates
Constructor Detail |
---|
public InfiniteMDP(States<S> initial)
initial
- set of initial states for the exploration
algorithmMethod Detail |
---|
public abstract Actions<A> feasibleActions(S i)
i
- Current State
public final int getNumStates()
protected abstract StatesSet<S> generate()
protected void setInterestRate(double interestRate)
interestRate
- effective interest rateprotected AbstractDiscountedSolver<S,A> getDefaultDiscountedSolver(double interestRate)
MDP.getDefaultSolver()
protected AbstractAverageSolver<S,A> getDefaultAverageSolver()
MDP.getDefaultSolver()
protected AbstractInfiniteSolver<S,A> getDefaultSolver()
MDP
getDefaultSolver
in class MDP<S extends State,A extends Action>
public StatesSet<S> getAllStates()
public AbstractInfiniteSolver<S,A> getSolver()
getSolver
in class MDP<S extends State,A extends Action>
|
||||||||||
PREV CLASS NEXT CLASS | FRAMES NO FRAMES | |||||||||
SUMMARY: NESTED | FIELD | CONSTR | METHOD | DETAIL: FIELD | CONSTR | METHOD |