Hardness Sampling for Self-Training Based Transductive Zero-Shot Learning.
Bo LiuQiulei DongZhanyi HuPublished in: CoRR (2021)
Keyphrases
- semi supervised learning
- semi supervised
- semi supervised classification
- co training
- knowledge transfer
- parameter space
- np complete
- sampling strategy
- computational complexity
- transductive learning
- training set
- unlabeled data
- phase transition
- random sampling
- image search
- sampling strategies
- monte carlo
- text classification
- np hard
- information theoretic
- sample size
- support vector
- machine learning
- data sets
- cost sensitive
- labeled data
- markov chain
- single view
- supervised learning
- kernel matrix
- sampling methods
- active learning