Global Neighbor Sampling for Mixed CPU-GPU Training on Giant Graphs.
Jialin DongDa ZhengLin F. YangGeorge KarypisPublished in: KDD (2021)
Keyphrases
- graphics processors
- gpu implementation
- training set
- graphics processing units
- supervised learning
- training examples
- graph theoretic
- graph databases
- data transfer
- training samples
- real time
- nearest neighbor
- training phase
- data sets
- bootstrap sampling
- graphics hardware
- uniform random
- graph representation
- graph mining
- bipartite graph
- connected components
- sample size
- back propagation
- online learning
- artificial neural networks