High-Performance Recommender System Training Using Co-Clustering on CPU/GPU Clusters.
Kubilay AtasuThomas P. ParnellCelestine DünnerMichail VlachosHaralampos PozidisPublished in: ICPP (2017)
Keyphrases
- graphics processing units
- recommender systems
- clustering algorithm
- collaborative filtering
- general purpose
- graphics processors
- gpu implementation
- graphics hardware
- parallel processing
- clustering framework
- heterogeneous computing
- matrix factorization
- document clustering
- high dimensional datasets
- graph clustering
- pc cluster
- parallel computing
- simultaneous clustering
- real time
- information filtering
- parallel implementation
- computing systems
- categorical data
- cluster analysis
- training process
- data sets
- training set
- compute intensive
- text mining
- training samples
- information theoretic
- data clustering
- high performance computing
- hierarchical clustering
- user profiles
- parallel computation
- product recommendation
- data points
- high dimensional
- data transfer