Scalable Batch Acquisition for Deep Bayesian Active Learning.
Aleksandr RubashevskiiDaria KotovaMaxim PanovPublished in: SDM (2023)
Keyphrases
- active learning
- batch mode
- belief nets
- incremental learning
- lightweight
- learning strategies
- maximum likelihood
- semi supervised
- bayesian inference
- bayesian learning
- deep learning
- bayesian networks
- experimental design
- imbalanced data classification
- selective sampling
- active learning strategies
- bayesian decision
- data driven
- supervised learning
- machine learning
- unlabeled data
- labeled data
- training examples
- pool based active learning
- highly scalable
- web scale
- relevance feedback
- sample selection
- real time
- neural network
- learning algorithm
- e learning
- training set
- online algorithms
- memory efficient
- acquisition process
- online learning
- cost sensitive