arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2102.03183cs.LGmath.OCstat.ML
收藏

最小二乘插值算法SGD的最后收敛性

Last iterate convergence of SGD for Least-Squares in the Interpolation regime

EPFL(洛桑联邦理工学院)

Aditya Varre, Loucas Pillaud-Vivien, Nicolas Flammarion

摘要:

受最近成功的神经网络,有能力完美地拟合数据和推广,我们研究了无噪声模型的基本最小二乘设置。我们假设最佳预测器完美地拟合输入和输出$\langle θ_* , ϕ(X) \rangle = Y$,其中$ϕ(X)$代表可能无限维的非线性特征映射。为了解决这个问题,我们考虑了由恒定步长的随机梯度下降(SGD)的最后一阶给出的估计。在这方面,我们的贡献有两个方面:(一)A(随机)优化的角度来看,我们展示了一个原型问题,我们可以显式地显示收敛的SGD最终收敛的非强凸问题与恒定的步长,而通常的结果使用某种形式的平均值和(ii)从统计的角度来看,我们在过参数化设置中给出了显式的非渐近收敛速率,并利用问题的细粒度参数化来展示可以比$O(1/T)$更快的多项式速率。建立了与再生核Hilbert空间的联系。

英文摘要:

Motivated by the recent successes of neural networks that have the ability to fit the data perfectly and generalize well, we study the noiseless model in the fundamental least-squares setup. We assume that an optimum predictor fits perfectly inputs and outputs $\langle θ_* , ϕ(X) \rangle = Y$, where $ϕ(X)$ stands for a possibly infinite dimensional non-linear feature map. To solve this problem, we consider the estimator given by the last iterate of stochastic gradient descent (SGD) with constant step-size. In this context, our contribution is two fold: (i) from a (stochastic) optimization perspective, we exhibit an archetypal problem where we can show explicitly the convergence of SGD final iterate for a non-strongly convex problem with constant step-size whereas usual results use some form of average and (ii) from a statistical perspective, we give explicit non-asymptotic convergence rates in the over-parameterized setting and leverage a fine-grained parameterization of the problem to exhibit polynomial rates that can be faster than $O(1/T)$. The link with reproducing kernel Hilbert spaces is established.

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。

↑