GrOD: Deep Learning with Gradients Orthogonal Decomposition for Knowledge Transfer, Distillation, and Adversarial Training.
Haoyi XiongRuosi WanJian ZhaoZeyu ChenXingjian LiZhanxing ZhuJun HuanPublished in: ACM Trans. Knowl. Discov. Data (2022)
Keyphrases
- knowledge transfer
- deep learning
- deep architectures
- restricted boltzmann machine
- knowledge sharing
- unsupervised feature learning
- transfer learning
- unsupervised learning
- machine learning
- supervised learning
- learning tasks
- training set
- weakly supervised
- mental models
- labeled data
- learning experience
- information sharing
- training examples
- training samples
- text mining
- semi supervised
- decision making
- information systems
- information retrieval