PSL: Exploiting Parallelism, Sparsity and Locality to Accelerate Matrix Factorization on x86 Platforms.
Weixin DengPengyu WangJing WangChao LiMinyi GuoPublished in: Bench (2019)
Keyphrases
- matrix factorization
- data sparsity
- cross domain collaborative filtering
- tensor factorization
- collaborative filtering
- recommender systems
- negative matrix factorization
- low rank
- missing data
- nonnegative matrix factorization
- sparsity constraints
- factorization methods
- stochastic gradient descent
- factor analysis
- data matrix
- binary matrix
- probabilistic matrix factorization
- sparse representation
- parallel computing
- factorization method
- signal processing
- variational bayesian
- item recommendation
- personalized ranking