In the misspecified spectral algorithms problem, researchers usually assume the underground true function $f_{\rho}^{*} \in [\mathcal{H}]^{s}$, a less-smooth interpolation space of a reproducing kernel Hilbert space (RKHS) $\mathcal{H}$ for some $s\in (0,1)$. The existing minimax optimal results require $\|f_{\rho}^{*}\|_{L^{\infty}}<\infty$ which implicitly requires $s > \alpha_{0}$ where $\alpha_{0}\in (0,1)$ is the embedding index, a constant depending on $\mathcal{H}$. Whether the spectral algorithms are optimal for all $s\in (0,1)$ is an outstanding problem lasting for years. In this paper, we show that spectral algorithms are minimax optimal for any $\alpha_{0}-\frac{1}{\beta} < s < 1$, where $\beta$ is the eigenvalue decay rate of $\mathcal{H}$. We also give several classes of RKHSs whose embedding index satisfies $ \alpha_0 = \frac{1}{\beta} $. Thus, the spectral algorithms are minimax optimal for all $s\in (0,1)$ on these RKHSs.
翻译:在误定谱算法问题中,研究人员通常假设底层真实函数 $f_{\rho}^{*} \in [\mathcal{H}]^{s}$,其中 $[\mathcal{H}]^{s}$ 是再生核希尔伯特空间(RKHS)$\mathcal{H}$ 的一个较低光滑度的插值空间,参数 $s\in (0,1)$。现有极小极大最优结果要求 $\|f_{\rho}^{*}\|_{L^{\infty}}<\infty$,这隐含地要求 $s > \alpha_{0}$,其中 $\alpha_{0}\in (0,1)$ 是嵌入指数,一个依赖于 $\mathcal{H}$ 的常数。谱算法是否对所有 $s\in (0,1)$ 都是最优的,是一个持续多年的未解问题。本文证明,对于任意满足 $\alpha_{0}-\frac{1}{\beta} < s < 1$ 的 $s$(其中 $\beta$ 是 $\mathcal{H}$ 的特征值衰减率),谱算法都是极小极大最优的。我们还给出了几类RKHS,其嵌入指数满足 $\alpha_0 = \frac{1}{\beta}$。因此,在这些RKHS上,谱算法对所有 $s\in (0,1)$ 都是极小极大最优的。