arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1612.00547math.OCcs.DS
收藏

梯度下降法求三次正则化非凸牛顿步

Gradient Descent Finds the Cubic-Regularized Non-Convex Newton Step

Yair Carmon, John C. Duchi

摘要:

考虑了三次项正则化的非凸二次型的极小化问题,该问题具有多个鞍点和局部极小点。尽管如此,我们证明,在温和的假设下,梯度下降近似的$\textit{global minimum}$内的$\varepsilon$的准确性在$O(\varepsilon^{-1}\log(1/\varepsilon))$步骤大$\varepsilon$和$O(\log(1/\varepsilon))$步骤小$\varepsilon$(相比,我们定义的条件数),最多对数依赖于问题的维度。当我们用梯度下降法来逼近正则化牛顿步时,我们的结果暗示了对一般光滑非凸函数的二阶稳定点的一个收敛速率。

英文摘要:

We consider the minimization of non-convex quadratic forms regularized by a cubic term, which exhibit multiple saddle points and poor local minima. Nonetheless, we prove that, under mild assumptions, gradient descent approximates the $\textit{global minimum}$ to within $\varepsilon$ accuracy in $O(\varepsilon^{-1}\log(1/\varepsilon))$ steps for large $\varepsilon$ and $O(\log(1/\varepsilon))$ steps for small $\varepsilon$ (compared to a condition number we define), with at most logarithmic dependence on the problem dimension. When we use gradient descent to approximate the cubic-regularized Newton step, our result implies a rate of convergence to second-order stationary points of general smooth non-convex functions.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑