Noisy Linear Convergence of Stochastic Gradient Descent for CV@R Statistical Learning under Polyak-Łojasiewicz Conditions.
Dionysios S. KalogeriasPublished in: CoRR (2020)
Keyphrases
- statistical learning
- stochastic gradient descent
- step size
- least squares
- weight vector
- model selection
- matrix factorization
- loss function
- supervised learning
- convergence rate
- information theory
- convergence speed
- maximum likelihood
- random forests
- support vector machine
- computer vision
- linear svm
- data sets
- manifold learning
- text mining
- active learning
- lower bound
- importance sampling
- machine learning