Emergence of heavy tails in homogenized stochastic gradient descent.
Zhe JiaoMartin Keller-ResselPublished in: CoRR (2024)
Keyphrases
- stochastic gradient descent
- heavy tails
- loss function
- least squares
- matrix factorization
- heavy tailed
- step size
- probability density function
- random forests
- support vector machine
- multiple kernel learning
- regularization parameter
- weight vector
- linear combination
- online algorithms
- image restoration
- image reconstruction
- support vector
- probability distribution
- multiresolution