ICML 2018oral54 citations

Stochastic Variance-Reduced Cubic Regularized Newton Methods

Dongruo Zhou, Pan Xu, Quanquan Gu

Abstract

We propose a stochastic variance-reduced cubic regularized Newton method (SVRC) for non-convex optimization. At the core of our algorithm is a novel semi-stochastic gradient along with a semi-stochastic Hessian, which are specifically designed for cubic regularization method. We show that our algorithm is guaranteed to converge to an $(\epsilon,\sqrt{\epsilon})$-approximate local minimum within $\tilde{O}(n^{4/5}/\epsilon^{3/2})$ second-order oracle calls, which outperforms the state-of-the-art cubic regularization algorithms including subsampled cubic regularization. Our work also sheds light on the application of variance reduction technique to high-order non-convex optimization methods. Thorough experiments on various non-convex optimization problems support our theory.

BibTeX
@InProceedings{pmlr-v80-zhou18d,
  title = 	 {Stochastic Variance-Reduced Cubic Regularized {N}ewton Methods},
  author =       {Zhou, Dongruo and Xu, Pan and Gu, Quanquan},
  booktitle = 	 {Proceedings of the 35th International Conference on Machine Learning},
  pages = 	 {5990--5999},
  year = 	 {2018},
  editor = 	 {Dy, Jennifer and Krause, Andreas},
  volume = 	 {80},
  series = 	 {Proceedings of Machine Learning Research},
  month = 	 {10--15 Jul},
  publisher =    {PMLR},
  pdf = 	 {http://proceedings.mlr.press/v80/zhou18d/zhou18d.pdf},
  url = 	 {https://proceedings.mlr.press/v80/zhou18d.html},
  abstract = 	 {We propose a stochastic variance-reduced cubic regularized Newton method (SVRC) for non-convex optimization. At the core of our algorithm is a novel semi-stochastic gradient along with a semi-stochastic Hessian, which are specifically designed for cubic regularization method. We show that our algorithm is guaranteed to converge to an $(\epsilon,\sqrt{\epsilon})$-approximate local minimum within $\tilde{O}(n^{4/5}/\epsilon^{3/2})$ second-order oracle calls, which outperforms the state-of-the-art cubic regularization algorithms including subsampled cubic regularization. Our work also sheds light on the application of variance reduction technique to high-order non-convex optimization methods. Thorough experiments on various non-convex optimization problems support our theory.}
}
Stochastic Variance-Reduced Cubic Regularized Newton Methods · ICML 2018