GST: Group-Sparse Training for Accelerating Deep Reinforcement Learning.
Juhyoung LeeSangyeob KimSangjin KimWooyoung JoHoi-Jun YooPublished in: CoRR (2021)
Keyphrases
- reinforcement learning
- function approximation
- supervised learning
- deep architectures
- markov decision processes
- learning process
- avoid overfitting
- training set
- high dimensional
- training algorithm
- transfer learning
- training samples
- sparse data
- temporal difference
- training process
- sparse representation
- training examples
- data sets
- state space
- multi agent
- training data
- machine learning
- neural network