Towards Understanding the True Loss Surface of Deep Neural Networks Using Random Matrix Theory and Iterative Spectral Methods

Cite

Text

Granziol et al. "Towards Understanding the True Loss Surface of Deep Neural Networks Using Random Matrix Theory and Iterative Spectral Methods." International Conference on Learning Representations, 2020.

Markdown

[Granziol et al. "Towards Understanding the True Loss Surface of Deep Neural Networks Using Random Matrix Theory and Iterative Spectral Methods." International Conference on Learning Representations, 2020.](https://mlanthology.org/iclr/2020/granziol2020iclr-understanding/)

BibTeX

@inproceedings{granziol2020iclr-understanding,
  title     = {{Towards Understanding the True Loss Surface of Deep Neural Networks Using Random Matrix Theory and Iterative Spectral Methods}},
  author    = {Granziol, Diego and Garipov, Timur and Vetrov, Dmitry and Zohren, Stefan and Roberts, Stephen and Wilson, Andrew Gordon},
  booktitle = {International Conference on Learning Representations},
  year      = {2020},
  url       = {https://mlanthology.org/iclr/2020/granziol2020iclr-understanding/}
}