跳到正文
返回论文列表
stat.ML提交于 已译

大维度下谱算法正则化不足的学习曲线与预测持续性

Learning Curves and Prediction Persistence to Under-Regularization of Spectral Algorithms in Large Dimensions

Weihao Lu · Qian Lin · Yingcun Xia · Dongming Huang

中文摘要

研究当样本量 n 与维度 d 满足 n≍d^γ(γ>0 固定)时,谱算法在正则化减弱过程中预测风险的变化。针对球面上的内积核,在光滑度 s≥0 的源条件和信号非退化条件下,刻画每个固定容许尺度 λ≍d^{-u}(u>0)及插值处的条件超额风险(至常数因子)。所得风险阶数描述源光滑度、滤波器资格与谱块间过渡的影响,识别从最优尺度减弱正则化至插值时条件超额风险何时保持极小极大风险阶,以及插值前某区间出现多项式更大风险的情形。还比较同一谱滤波器的核风险与总体序列风险,采用下界为 n^{-1} 阶的确定性正则化尺度。最后,对非整数 γ,在谱缩放、超收缩性与尾部对角集中条件下,结合源与信号条件,将核岭回归的风险刻画推广到一般域上的核。

关键要点

  1. 01研究 n≍d^γ 时谱算法正则化减弱下的预测风险变化
  2. 02刻画球面内积核在源条件与信号非退化下的条件超额风险阶
  3. 03识别正则化减弱至插值时风险保持极小极大阶或多项式增大的条件
  4. 04比较核风险与总体序列风险,正则化尺度下界为 n^{-1} 阶
  5. 05非整数 γ 下将核岭回归风险刻画推广到一般域核

解读

尚无解读。

原始英文摘要

arXiv:2604.23212v2 Announce Type: replace Abstract: We study how prediction risk changes as regularization is reduced for spectral algorithms when the sample size $n$ and dimension $d$ satisfy $n\asymp d^\gamma$ for fixed $\gamma>0$. For inner-product kernels on the sphere, under a source condition with smoothness $s\geq0$ and a signal nondegeneracy condition, we characterize the conditional excess risk up to constant factors at each fixed admissible scale $\lambda\asymp d^{-u}$, $u>0$, and at interpolation. The resulting risk orders describe the effects of source smoothness, filter qualification, and transitions between spectral blocks. They identify when the conditional excess risk retains the minimax risk order as regularization is weakened from an optimal scale to interpolation, and when an interval before interpolation exhibits a polynomially larger risk. We also compare the kernel risk with a population sequence risk for the same spectral filter, using a deterministic regularization scale bounded below by order $n^{-1}$. Finally, for noninteger $\gamma$, we extend the risk characterization for kernel ridge regression to kernels on general domains under spectral scaling, hyper-contractivity, and tail diagonal concentration conditions, together with the source and signal conditions.

同方向论文 · stat.ML

查看全部 →