arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2202.04005cs.LGstat.ML
收藏

基于核学习的稀疏近似方法的收敛速度改进

Improved Convergence Rates for Sparse Approximation Methods in Kernel-Based Learning

MediaTek Research(联发科研究院);National University of Singapore(新加坡国立大学)

Sattar Vakili, Jonathan Scarlett, Da-shan Shiu, Alberto Bernacchia

摘要:

核岭回归和高斯过程等基于核的模型在用于回归和优化的机器学习应用中是普遍存在的。众所周知,基于内核的模型的一个主要缺点是计算成本高;给定$n$样本的数据集,成本随着$\mathcal{O}(n^3)$而增长。现有的稀疏近似方法可以显著降低计算成本,在某些情况下有效地将实际成本降低到低至$\mathcal{O}(n)$。尽管这一显着的经验成功,显着的差距仍然存在于现有的结果的分析界限的误差,由于近似。在这项工作中,我们提供了新的置信区间的Nyström方法和稀疏变分高斯过程近似方法,我们建立使用新的解释近似(替代)后验方差的模型。我们的置信区间导致在回归和优化问题的性能界限改善。

英文摘要:

Kernel-based models such as kernel ridge regression and Gaussian processes are ubiquitous in machine learning applications for regression and optimization. It is well known that a major downside for kernel-based models is the high computational cost; given a dataset of $n$ samples, the cost grows as $\mathcal{O}(n^3)$. Existing sparse approximation methods can yield a significant reduction in the computational cost, effectively reducing the actual cost down to as low as $\mathcal{O}(n)$ in certain cases. Despite this remarkable empirical success, significant gaps remain in the existing results for the analytical bounds on the error due to approximation. In this work, we provide novel confidence intervals for the Nyström method and the sparse variational Gaussian process approximation method, which we establish using novel interpretations of the approximate (surrogate) posterior variance of the models. Our confidence intervals lead to improved performance bounds in both regression and optimization problems.

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。

↑