Utilizing Half Convolutional Autoencoder to Generate User and Item Vectors for Initialization in Matrix Factorization.
Tan Nghia DuongNguyen Nam DoanTruong Giang DoManh Hoang TranDuc Minh NguyenQuang Hieu DangPublished in: Future Internet (2022)
Keyphrases
- matrix factorization
- item recommendation
- collaborative filtering
- recommender systems
- cold start problem
- personalized ranking
- low rank
- user ratings
- user preferences
- nonnegative matrix factorization
- tag recommendation
- missing data
- tensor factorization
- latent factors
- implicit feedback
- factor analysis
- data sparsity
- negative matrix factorization
- factorization methods
- latent factor models
- data matrix
- user interaction
- stochastic gradient descent
- cold start
- binary matrix
- probabilistic matrix factorization
- higher order
- user interface
- personalized recommendation
- image classification
- least squares
- feature vectors