def __init__(self, obs_noise, init_true_state, init_belief):
"""init_belief is a Distribution."""
agent = pomdp_py.Agent(init_belief, PolicyModel(), TransitionModel(),
ObservationModel(obs_noise), RewardModel())
env = pomdp_py.Environment(init_true_state, TransitionModel(),
RewardModel())
super().__init__(agent, env, name="TigerProblem")
def __init__(self, init_state, init_belief):
"""init_belief is a Distribution."""
agent = pomdp_py.Agent(init_belief, LUPolicyModel(),
LUTransitionModel(), LUObservationModel(),
LURewardModel())
env = pomdp_py.Environment(init_state, LUTransitionModel(),
LURewardModel())
super().__init__(agent, env, name="LoadUnloadProblem")
def __init__(self, n, k, init_state, rock_locs, init_belief):
self._n, self._k = n, k
agent = pomdp_py.Agent(
init_belief, RSPolicyModel(n, k),
RSTransitionModel(n, rock_locs, self.in_exit_area),
RSObservationModel(rock_locs),
RSRewardModel(rock_locs, self.in_exit_area))
env = pomdp_py.Environment(
init_state, RSTransitionModel(n, rock_locs, self.in_exit_area),
RSRewardModel(rock_locs, self.in_exit_area))
self._rock_locs = rock_locs
super().__init__(agent, env, name="RockSampleProblem")