Incremental and Approximate Inference for Faster Occlusion-based Deep CNN Explanations.
Supun NakandalaArun KumarYannis PapakonstantinouPublished in: SIGMOD Conference (2019)
Keyphrases
- approximate inference
- graphical models
- belief propagation
- probabilistic inference
- exact inference
- gaussian process
- message passing
- bayesian networks
- parameter estimation
- factor graphs
- conditional random fields
- loopy belief propagation
- variational methods
- latent variables
- deep belief networks
- expectation propagation
- dynamic bayesian networks
- structured prediction
- incremental learning
- free energy
- random variables
- dynamic programming
- image processing
- markov random field
- conditional probabilities
- generalized belief propagation