BayesCap: Bayesian Identity Cap for Calibrated Uncertainty in Frozen Neural Networks.
Uddeshya UpadhyayShyamgopal KarthikYanbei ChenMassimiliano ManciniZeynep AkataPublished in: CoRR (2022)
Keyphrases
- neural network
- decision theory
- dempster shafer
- pattern recognition
- artificial neural networks
- back propagation
- uncertain data
- maximum likelihood
- bayesian decision
- bayesian networks
- expected utility
- bayesian learning
- neural network model
- fuzzy logic
- inherent uncertainty
- conditional probabilities
- social context
- bayesian analysis
- neural nets
- bayesian inference
- posterior distribution
- network architecture
- bayesian estimation
- rule extraction
- multilayer perceptron
- recurrent neural networks
- posterior probability
- incomplete information
- multi view
- data driven
- rough sets
- decision making
- genetic algorithm