Feature Learning in $L_2$-regularized DNNs: Attraction/Repulsion and Sparsity.
Arthur JacotEugene GolikovClément HonglerFranck GabrielPublished in: NeurIPS (2022)
Keyphrases
- learning algorithm
- learning process
- reinforcement learning
- learning systems
- online learning
- prior knowledge
- attraction repulsion
- learning problems
- case study
- genetic algorithm
- feature space
- training data
- active learning
- mobile robot
- feature selection
- knowledge acquisition
- mobile learning
- learning tasks
- machine learning
- learning community
- data sets