On the instability of embeddings for recommender systems: the case of Matrix Factorization.
Giovanni GabboliniEdoardo D'AmicoCesare BernardisPaolo CremonesiPublished in: CoRR (2021)
Keyphrases
- matrix factorization
- recommender systems
- collaborative filtering
- low rank
- cold start problem
- data sparsity
- nonnegative matrix factorization
- factorization methods
- user preferences
- implicit feedback
- negative matrix factorization
- missing data
- item recommendation
- probabilistic matrix factorization
- factor analysis
- user profiles
- stochastic gradient descent
- variational bayesian
- tensor factorization
- user interests
- factorization method
- recommendation systems
- latent space
- feature extraction
- data representation
- linear combination
- low dimensional
- high dimensional