Global Convergence of SGD For Logistic Loss on Two Layer Neural Nets
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2023
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917615862022144 |
|---|---|
| author | Gopalani, Pulkit Jha, Samyak Mukherjee, Anirbit |
| author_facet | Gopalani, Pulkit Jha, Samyak Mukherjee, Anirbit |
| contents | In this note, we demonstrate a first-of-its-kind provable convergence of SGD to the global minima of appropriately regularized logistic empirical risk of depth $2$ nets -- for arbitrary data and with any number of gates with adequately smooth and bounded activations like sigmoid and tanh. We also prove an exponentially fast convergence rate for continuous time SGD that also applies to smooth unbounded activations like SoftPlus. Our key idea is to show the existence of Frobenius norm regularized logistic loss functions on constant-sized neural nets which are "Villani functions" and thus be able to build on recent progress with analyzing SGD on such objectives. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2309_09258 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | Global Convergence of SGD For Logistic Loss on Two Layer Neural Nets Gopalani, Pulkit Jha, Samyak Mukherjee, Anirbit Machine Learning Optimization and Control In this note, we demonstrate a first-of-its-kind provable convergence of SGD to the global minima of appropriately regularized logistic empirical risk of depth $2$ nets -- for arbitrary data and with any number of gates with adequately smooth and bounded activations like sigmoid and tanh. We also prove an exponentially fast convergence rate for continuous time SGD that also applies to smooth unbounded activations like SoftPlus. Our key idea is to show the existence of Frobenius norm regularized logistic loss functions on constant-sized neural nets which are "Villani functions" and thus be able to build on recent progress with analyzing SGD on such objectives. |
| title | Global Convergence of SGD For Logistic Loss on Two Layer Neural Nets |
| topic | Machine Learning Optimization and Control |
| url | https://arxiv.org/abs/2309.09258 |