Partial Policy-Based Reinforcement Learning for Anatomical Landmark Localization in 3D Medical Images.
Walid Abdullah AlIl Dong YunPublished in: IEEE Trans. Medical Imaging (2020)
Keyphrases
- landmark localization
- reinforcement learning
- optimal policy
- policy search
- markov decision process
- face detection
- action selection
- reinforcement learning algorithms
- three dimensional
- actor critic
- markov decision processes
- reinforcement learning problems
- partially observable environments
- approximate dynamic programming
- function approximators
- medical images
- state space
- action space
- partially observable
- function approximation
- reward function
- control policy
- markov decision problems
- control policies
- agent learns
- state action
- policy evaluation
- dynamic programming
- inverse reinforcement learning
- state and action spaces
- average reward
- partially observable domains
- continuous state spaces
- exploration exploitation tradeoff
- policy gradient
- reinforcement learning methods
- partially observable markov decision processes
- temporal difference
- model free
- anatomical structures
- decision problems
- rl algorithms
- policy iteration
- finite state
- optimal control
- agent receives
- state dependent
- x ray
- policy gradient methods
- learning algorithm