Learning sparse features can lead to overfitting in neural networks.
Leonardo PetriniFrancesco CagnettaEric Vanden-EijndenMatthieu WyartPublished in: NeurIPS (2022)
Keyphrases
- neural network
- learning process
- learning algorithm
- reinforcement learning
- feature vectors
- prior knowledge
- neural nets
- online learning
- cross validation
- connectionist systems
- error back propagation
- multi layer
- learning models
- incremental learning
- learning problems
- knowledge acquisition
- supervised learning
- image features
- pattern recognition
- unsupervised learning
- linear combination
- back propagation
- learning systems
- co occurrence
- active learning
- high dimensional
- feedforward neural networks
- cascade correlation
- decision trees