Deep Distributional Reinforcement Learning Based High-Level Driving Policy Determination.

Kyushik Min Hayoung Kim Kunsoo Huh

Published in: IEEE Trans. Intell. Veh. (2019)

Keyphrases

reinforcement learning
high level
optimal policy
policy search
low level
action selection
markov decision process
function approximators
reward function
actor critic
state and action spaces
function approximation
partially observable environments
control policies

policy gradient
markov decision processes
state action
action space
markov decision problems
partially observable
higher level
co occurrence
policy iteration
decision problems
state space
control policy
approximate dynamic programming
programming language

reinforcement learning algorithms
lower level
reinforcement learning problems
dynamic programming
partially observable markov decision processes
temporal difference
model free reinforcement learning
agent learns
policy evaluation
deep learning
learning algorithm
finite state
multi agent

low level features
sufficient conditions
policy gradient methods
partially observable domains
source code
neural network
rl algorithms
transition model
optimal control
infinite horizon
model free
continuous state
average cost