Overall error analysis for the training of deep neural networks via stochastic gradient descent with random initialisation.
Arnulf JentzenTimo WeltiPublished in: CoRR (2020)
Keyphrases
- error analysis
- stochastic gradient descent
- least squares
- neural network
- early stopping
- training speed
- loss function
- step size
- matrix factorization
- training process
- error correction
- random forests
- training algorithm
- regularization parameter
- weight vector
- online algorithms
- supervised learning
- back propagation
- collaborative filtering
- importance sampling
- support vector machine
- three dimensional