Phase transitions and optimal algorithms in the semi-supervised classfications in graphs: from belief propagation to convolution neural networks.
Pengfei ZhouPan ZhangPublished in: CoRR (2019)
Keyphrases
- belief propagation
- neural network
- phase transition
- message passing
- graphical models
- loopy belief propagation
- graph cuts
- max product
- stereo matching
- markov random field
- semi supervised
- learning algorithm
- dynamic programming
- pairwise
- multi view
- approximate inference
- hard problems
- upper bound
- fixed point
- np hard
- random graphs
- image sequences