Toward Optimal Probabilistic Active Learning Using a Bayesian Approach.
Daniel KottkeMarek HerdeChristoph SandrockDenis HuseljicGeorg KremplBernhard SickPublished in: CoRR (2020)
Keyphrases
- active learning
- probabilistic model
- optimal design
- selective sampling
- generative model
- dynamic programming
- random sampling
- semi supervised
- relevance feedback
- labeled data
- clustering algorithm
- state space
- database
- worst case
- graphical models
- training set
- search algorithm
- learning strategies
- uncertain data
- experimental design
- information retrieval
- data sets