DePRL: Achieving Linear Convergence Speedup in Personalized Decentralized Learning with Shared Representations.
Guojun XiongGang YanShiqiang WangJian LiPublished in: CoRR (2023)
Keyphrases
- learning process
- learning algorithm
- learning systems
- inductive inference
- learning tasks
- reinforcement learning
- multi agent
- learning scheme
- artificial neural networks
- adaptive learning
- peer to peer
- knowledge acquisition
- learning problems
- external representations
- background knowledge
- orders of magnitude
- supervised learning
- object recognition
- training data
- knowledge base