Accelerating Stochastic Gradient Descent via Online Learning to Sample.
Guillaume BouchardThéo TrouillonJulien PerezAdrien GaidonPublished in: CoRR (2015)
Keyphrases
- online learning
- stochastic gradient descent
- online algorithms
- loss function
- least squares
- matrix factorization
- e learning
- random forests
- step size
- sample size
- multiple kernel learning
- feature extraction
- weight vector
- regularization parameter
- pairwise
- support vector machine
- image restoration
- small number
- active learning
- importance sampling
- lower bound
- learning process