Learn Locally, Correct Globally: A Distributed Algorithm for Training Graph Neural Networks.
Morteza RamezaniWeilin CongMehrdad MahdaviMahmut T. KandemirAnand SivasubramaniamPublished in: ICLR (2022)
Keyphrases
- neural network
- dynamic programming
- training algorithm
- training process
- preprocessing
- learning algorithm
- k means
- cost function
- simulated annealing
- clustering method
- training phase
- neural network training
- graph based algorithm
- feed forward neural networks
- graph structure
- detection algorithm
- expectation maximization
- worst case
- similarity measure
- matching algorithm
- tree structure
- graph databases
- globally optimal
- multi agent
- objective function