Dynamic Sparse Training via Balancing the Exploration-Exploitation Trade-off.
Shaoyi HuangBowen LeiDongkuan XuHongwu PengYue SunMimi XieCaiwen DingPublished in: CoRR (2022)
Keyphrases
- dynamic environments
- dynamically changing
- training process
- online learning
- knowledge base
- high dimensional
- bayesian networks
- neural network
- training examples
- supervised learning
- case study
- data structure
- decision trees
- information systems
- artificial intelligence
- learning algorithm
- training samples
- machine learning
- test set
- sparse representation
- real world
- random projections
- database
- avoid overfitting