Quantifying the generalization error in deep learning in terms of data distribution and neural network smoothness.
Pengzhan JinLu LuYifa TangGeorge E. KarniadakisPublished in: CoRR (2019)
Keyphrases
- data distribution
- deep learning
- generalization error
- neural network
- data streams
- upper bound
- machine learning
- classification error
- active learning
- unsupervised learning
- model selection
- learning algorithm
- sample size
- high dimensional data
- data points
- cross validation
- training data
- supervised learning
- artificial neural networks
- index structure
- weakly supervised
- database
- neural nets
- data mining
- data sets
- worst case
- pairwise
- lower bound
- pattern recognition
- objective function