MDL4BMF: Minimum Description Length for Boolean Matrix Factorization.
Pauli MiettinenJilles VreekenPublished in: ACM Trans. Knowl. Discov. Data (2014)
Keyphrases
- minimum description length
- matrix factorization
- information theoretic
- mdl principle
- collaborative filtering
- recommender systems
- low rank
- binary matrix
- negative matrix factorization
- factor analysis
- factorization methods
- nonnegative matrix factorization
- missing data
- data sparsity
- data matrix
- information theory
- latent factors
- tensor factorization
- implicit feedback
- variational bayesian
- stochastic gradient descent
- least squares
- high dimensional
- item recommendation
- probabilistic matrix factorization