Probabilistically Rewired Message-Passing Neural Networks.
Chendi QianAndrei ManolacheKareem AhmedZhe ZengGuy Van den BroeckMathias NiepertChristopher MorrisPublished in: ICLR (2024)
Keyphrases
- message passing
- neural network
- belief propagation
- probabilistic inference
- distributed systems
- shared memory
- factor graphs
- markov random field
- approximate inference
- distributed shared memory
- sum product algorithm
- inference in graphical models
- back propagation
- graphical models
- object recognition
- graph cuts
- competitive learning
- loopy belief propagation
- artificial neural networks
- ldpc codes
- max product
- three dimensional
- computer vision