Directional Smoothness and Gradient Methods: Convergence and Adaptivity
We develop new sub-optimality bounds for gradient descent (GD) that depend on the conditioning of the objective along the path of optimization rather than on global, worst-case constants. Key to our proofs is directional smoothness, a measure of gradient variation that we use to develop upper-bounds...
שמור ב:
| Autores principales: | , , , , |
|---|---|
| פורמט: | Artículo |
| יצא לאור: |
2024
|
| נושאים: | |
| גישה מקוונת: | https://arxiv.org/abs/2403.04081 http://arxiv.org/abs/2403.04081 |
| תגים: |
אין תגיות, היה/י הראשונ/ה לתייג את הרשומה!
|