Scalability of Stochastic Gradient Descent based on "Smart" Sampling Techniques.
Stéphan ClémençonAurélien BelletOns JelassiGuillaume PapaPublished in: INNS Conference on Big Data (2015)
Keyphrases
- stochastic gradient descent
- importance sampling
- least squares
- loss function
- matrix factorization
- random forests
- step size
- regularization parameter
- monte carlo
- random sampling
- weight vector
- online algorithms
- optical flow
- multiple kernel learning
- support vector machine
- sample size
- markov chain monte carlo
- convergence speed
- markov chain
- principal component analysis