Bias Correction of Learned Generative Models using Likelihood-Free Importance Weighting.
Aditya GroverJiaming SongAlekh AgarwalKenneth TranAshish KapoorEric HorvitzStefano ErmonPublished in: CoRR (2019)
Keyphrases
- labeled data
- semi supervised
- bias correction
- generative model
- semi supervised learning
- posterior probability
- discriminative methods
- discriminative models
- maximum entropy principle
- naive bayes models
- active learning
- prior knowledge
- mixture model
- unsupervised learning
- bayesian framework
- discriminative learning
- maximum likelihood
- conditional random fields
- similarity measure
- decision trees
- support vector
- probabilistic model
- hidden variables
- topic models
- latent dirichlet allocation
- maximum entropy
- image features
- naive bayes
- bayesian networks
- higher order
- data sets