FedFAME: A Data Augmentation Free Framework based on Model Contrastive Learning for Federated Semi-Supervised Learning.
Shubham MalaviyaManish ShuklaPratik KoratSachin LodhaPublished in: SAC (2023)
Keyphrases
- learning models
- semi supervised learning
- prior knowledge
- unlabeled data
- labeled data
- boosting framework
- supervised learning
- learning frameworks
- labeled and unlabeled data
- semi supervised
- multi view learning
- data sets
- learning process
- test data
- data sources
- active learning
- learning problems
- manifold regularization
- probabilistic model
- unsupervised learning
- original data
- discriminative models
- learning tasks
- data analysis
- training data
- machine learning
- missing data
- data points
- learning algorithm
- spectral clustering
- transfer learning
- multi task
- metric learning
- co training
- data distribution
- training set
- partially labeled
- reinforcement learning
- similarity measure
- semi supervised learning setting