Flat Posterior Does Matter For Bayesian Transfer Learning.
Sungjun LimJeyoon YeomSooyon KimHoyoon ByunJinho KangYohan JungJiyoung JungKyungwoo SongPublished in: CoRR (2024)
Keyphrases
- transfer learning
- posterior distribution
- posterior probability
- learning tasks
- probability distribution
- knowledge transfer
- multi task
- reinforcement learning
- cross domain
- structure learning
- manifold alignment
- parameter estimation
- bayesian framework
- domain adaptation
- machine learning
- latent variables
- machine learning algorithms
- text categorization
- semi supervised learning
- labeled data
- active learning
- text classification
- multi task learning
- transfer knowledge
- transferring knowledge
- collaborative filtering
- maximum likelihood
- bayesian networks
- hyperparameters
- bayesian inference
- learning algorithm
- target domain
- knowledge discovery
- text mining
- knn
- information retrieval