Every Parameter Matters: Ensuring the Convergence of Federated Learning with Dynamic Heterogeneous Models Reduction.
Hanhan ZhouTian LanGuru VenkataramaniWenbo DingPublished in: NeurIPS (2023)
Keyphrases
- learning algorithm
- prior knowledge
- accurate models
- learning process
- learning models
- active learning
- dynamic environments
- learning tasks
- learned models
- supervised learning
- unsupervised learning
- mobile learning
- parameter estimation
- learning analytics
- learning problems
- statistical models
- machine learning algorithms
- state space
- probabilistic model
- training data