Speeding Up Budgeted Stochastic Gradient Descent SVM Training with Precomputed Golden Section Search.
Tobias GlasmachersSahar QaadanPublished in: CoRR (2018)
Keyphrases
- stochastic gradient descent
- svm training
- online algorithms
- loss function
- step size
- least squares
- generalization ability
- matrix factorization
- random forests
- support vector machine
- importance sampling
- multi class
- kernel methods
- cost function
- weight vector
- svm classification
- random forest
- support vectors
- average case
- support vector machine svm