Breaking the Limits of Message Passing Graph Neural Networks.
Muhammet BalcilarPierre HérouxBenoit GaüzèrePascal VasseurSébastien AdamPaul HoneinePublished in: ICML (2021)
Keyphrases
- message passing
- neural network
- belief propagation
- distributed systems
- shared memory
- inference in graphical models
- max product
- factor graphs
- probabilistic inference
- junction tree
- approximate inference
- tree reweighted
- markov random field
- sum product
- distributed shared memory
- back propagation
- singly connected
- artificial neural networks
- matrix multiplication
- random variables
- graph cuts
- ldpc codes
- graphical models