Verifying message-passing neural networks via topology-based bounds tightening.
Christopher HojnyShiqiang ZhangJuan S. CamposRuth MisenerPublished in: CoRR (2024)
Keyphrases
- message passing
- neural network
- belief propagation
- distributed systems
- probabilistic inference
- factor graphs
- shared memory
- upper bound
- artificial neural networks
- lower bound
- markov random field
- distributed shared memory
- back propagation
- approximate inference
- sum product
- sum product algorithm
- competitive learning
- inference in graphical models
- ldpc codes
- graphical models
- lower and upper bounds
- matrix multiplication
- message passing interface
- energy function
- max product
- pairwise
- preprocessing
- computer vision