Smaller generalization error derived for deep compared to shallow residual neural networks.
Aku KammonenJonas KiesslingPetr PlechácMattias SandbergAnders SzepessyRaúl TemponePublished in: CoRR (2020)
Keyphrases
- generalization error
- neural network
- learning algorithm
- active learning
- training error
- cross validation
- model selection
- sample size
- generalization error bounds
- upper bound
- fault diagnosis
- neural nets
- classification error
- binary classification
- semi supervised
- neural network model
- recurrent neural networks
- learning theory
- fuzzy logic
- linear classifiers
- machine learning