Active Learning for Approximation of Expensive Functions with Normal Distributed Output Uncertainty.
Joachim van der HertenIvo CouckuytDirk DeschrijverTom DhaenePublished in: CoRR (2016)
Keyphrases
- active learning
- distributed systems
- cooperative
- distributed environment
- multi agent
- peer to peer
- neural network
- continuous functions
- learning strategies
- mobile agents
- learning algorithm
- convex functions
- training examples
- lightweight
- supervised learning
- semi supervised
- computer networks
- error bounds
- error prone
- computing environments
- random sampling
- experimental design
- distributed data
- decision theory
- pairwise
- selective sampling