Train faster, generalize better: Stability of stochastic gradient descent.
Moritz HardtBenjamin RechtYoram SingerPublished in: CoRR (2015)
Keyphrases
- evolutionary algorithm
- stochastic gradient descent
- least squares
- matrix factorization
- loss function
- step size
- random forests
- multiple kernel learning
- importance sampling
- online algorithms
- support vector machine
- regularization parameter
- weight vector
- cost function
- collaborative filtering
- machine learning
- multiresolution
- learning algorithm