Auto-tuning HyperParameters of SGD Matrix Factorization-Based Recommender Systems Using Genetic Algorithm.
Habib IraniFatemeh ElahiMahmood FazlaliMahyar ShahsavariBahareh J. FarahaniPublished in: COINS (2022)
Keyphrases
- matrix factorization
- recommender systems
- hyperparameters
- stochastic gradient descent
- genetic algorithm
- model selection
- collaborative filtering
- cross validation
- closed form
- parameter settings
- bayesian inference
- random sampling
- support vector
- bayesian framework
- gaussian process
- noise level
- posterior distribution
- prior information
- maximum likelihood
- low rank
- em algorithm
- sample size
- maximum a posteriori
- gaussian processes
- incremental learning
- genetic algorithm ga
- regularization parameter
- negative matrix factorization
- nonnegative matrix factorization
- incomplete data
- missing values
- implicit feedback
- parameter space
- probabilistic matrix factorization
- missing data
- active learning
- expectation maximization
- data sparsity
- training set
- latent factor models
- latent factors
- data sets