PCA Initialization for Approximate Message Passing in Rotationally Invariant Models.
Marco MondelliRamji VenkataramananPublished in: NeurIPS (2021)
Keyphrases
- message passing
- inference in graphical models
- belief propagation
- rotationally invariant
- factor graphs
- graphical models
- probabilistic model
- principal component analysis
- shared memory
- sum product algorithm
- feature space
- pattern recognition
- junction tree
- pairwise
- distributed systems
- feature extraction
- approximate inference
- bayesian networks