Learning sparse features can lead to overfitting in neural networks.
Leonardo PetriniFrancesco CagnettaEric Vanden-EijndenMatthieu WyartPublished in: CoRR (2022)
Keyphrases
- neural network
- learning process
- learning tasks
- learning systems
- reinforcement learning
- neural nets
- prior knowledge
- active learning
- feature set
- image features
- multi layer
- dictionary learning
- learning rules
- feature space
- decision trees
- learning algorithm
- canonical correlation analysis
- activation function
- error back propagation
- learning models
- mobile learning
- knowledge acquisition
- online learning
- training set