Kernel partial least squares for stationary data

2017/06/12 by Marco Singer, Singer, Marco, Tatyana Krivobokova +3
Chemistry · Mathematics · Environmental Science · #Spectroscopy and Chemometric Analyses #Statistical and numerical algorithms #Soil Geostatistics and Mapping

paper · pdf · doi:10.48550/arxiv.1706.03559

Abstract

We consider the kernel partial least squares algorithm for non-parametric regression with stationary dependent data. Probabilistic convergence rates of the kernel partial least squares estimator to the true regression function are established under a source and an effective dimensionality condition. It is shown both theoretically and in simulations that long range dependence results in slower convergence rates. A protein dynamics example shows high predictive power of kernel partial least squares.

Citations

• Detection of Functional Modes in Protein Dynamics
• A Statistical View of Some Chemometrics Regression Tools

Related

• Partial least squares for dependent data
• Kernel regression, minimax rates and effective dimensionality: beyond\n the regular case
• Optimal learning rates for Kernel Conjugate Gradient regression
• Early stopping and non-parametric regression: An optimal data-dependent stopping rule
• Convergence Rates for Kernel Regression in Infinite Dimensional Spaces