Sparse Tensor Factorization on Many-Core Processors with High-Bandwidth Memory.
Shaden SmithJongsoo ParkGeorge KarypisPublished in: IPDPS (2017)
Keyphrases
- tensor factorization
- high bandwidth
- end to end
- matrix factorization
- negative matrix factorization
- tensor decomposition
- collaborative filtering
- parallel architectures
- low latency
- application specific
- gaussian processes
- linear algebra
- high density
- parallel processing
- parallel algorithm
- link prediction
- probabilistic latent semantic analysis
- singular value decomposition
- nonnegative matrix factorization
- factorization method
- tag recommendation
- recommender systems
- mobile terminals
- parallel computing
- parallel computers
- kl divergence
- parallel programming
- real time
- context aware
- high speed
- image processing