Personalized Federated Learning for Fast Convergence and Local Adaption: A Double Gradient Descent Method.
Zhidu LiSongyang HeXu LiMeiting XuePublished in: HPCC/DSS/SmartCity/DependSys (2023)
Keyphrases
- cost function
- learning scheme
- unsupervised learning
- significant improvement
- reinforcement learning
- learning process
- dynamic programming
- objective function
- preprocessing
- faster convergence
- learning mechanism
- learning systems
- high accuracy
- support vector machine
- probabilistic model
- prior knowledge
- image sequences
- learning algorithm
- machine learning
- k means
- classification accuracy
- pairwise
- feature set
- clustering algorithm
- learning styles
- convergence rate
- neural nets
- adaptive learning
- genetic algorithm
- neural network