MINDFL: Mitigating the Impact of Imbalanced and Noisy-labeled Data in Federated Learning with Quality and Fairness-Aware Client Selection.
Chaoyu ZhangNing WangShanghao ShiChanglai DuWenjing LouY. Thomas HouPublished in: MILCOM (2023)
Keyphrases
- labeled data
- supervised learning
- active learning
- learning algorithm
- semi supervised learning
- unlabeled data
- labeled and unlabeled data
- semi supervised
- domain adaptation
- prior knowledge
- sample selection bias
- training data
- co training
- learning tasks
- transfer learning
- learning problems
- supervised learning algorithms
- partially labeled data
- labeled instances
- machine learning
- text classification
- data points
- learning process
- reinforcement learning
- background knowledge
- unsupervised learning
- inductive inference
- labeled examples
- multi class
- semi supervised classification
- labeled training data
- labeling process
- transferring knowledge
- data sets