Normal Approximation for Stochastic Gradient Descent via Non-Asymptotic Rates of Martingale CLT.
Andreas AnastasiouKrishnakumar BalasubramanianMurat A. ErdogduPublished in: COLT (2019)
Keyphrases
- stochastic gradient descent
- loss function
- matrix factorization
- least squares
- step size
- importance sampling
- random forests
- online algorithms
- support vector machine
- closed form
- weight vector
- worst case
- regularization parameter
- multiple kernel learning
- asymptotically optimal
- collaborative filtering
- cost function
- logistic regression
- image sequences
- feature space