The Inductive Bias of Flatness Regularization for Deep Matrix Factorization.
Khashayar GatmiryZhiyuan LiChing-Yao ChuangSashank ReddiTengyu MaStefanie JegelkaPublished in: CoRR (2023)
Keyphrases
- matrix factorization
- inductive bias
- machine learning
- collaborative filtering
- inductive learning
- pac learning
- prior knowledge
- low rank
- supervised learning
- recommender systems
- domain knowledge
- domain theory
- training examples
- factorization methods
- hypothesis space
- missing data
- stochastic gradient descent
- multitask learning
- training set
- nonnegative matrix factorization
- decision trees
- active learning
- learning tasks
- machine learning algorithms
- learning theory
- training data
- learning problems
- training samples
- unsupervised learning
- sample selection
- learning algorithm