Rl-Ncs: Reinforcement Learning Based Data-Driven Approach For Nonuniform Compressed Sensing.
Nazmul KarimAlireza ZaeemzadehNazanin RahnavardPublished in: MLSP (2019)
Keyphrases
- compressed sensing
- reinforcement learning
- image reconstruction
- natural images
- sparse representation
- random projections
- function approximation
- model free
- rl algorithms
- reinforcement learning algorithms
- state space
- temporal difference
- compressive sampling
- markov decision processes
- machine learning
- supervised learning
- fourier domain
- learning algorithm
- optimal policy
- random sampling
- signal processing
- higher order
- action space
- fourier transform
- dynamic programming
- learning process
- image processing
- orthogonal matching pursuit