A matrix factorization approach to graph compression with partial information.
Farshad NourbakhshSamuel Rota BulòMarcello PelilloPublished in: Int. J. Mach. Learn. Cybern. (2015)
Keyphrases
- matrix factorization
- partial information
- incomplete information
- collaborative filtering
- recommender systems
- low rank
- missing data
- sparse non negative
- nonnegative matrix factorization
- negative matrix factorization
- factor analysis
- factorization methods
- data sparsity
- random walk
- stochastic gradient descent
- item recommendation
- binary matrix
- probabilistic matrix factorization
- singular value decomposition
- domain specific
- factor matrices
- pairwise