TOTAL VARIATION REGULARIZED FRECHET REGRESSION FOR METRIC-SPACE VALUED DATA

• 成果类型: Article
• 署名作者: Lin, Zhenhua; Muller, Hans-Georg
• 署名单位: National University of Singapore; University of California System; University of California Davis
• 刊物名称: ANNALS OF STATISTICS
• ISSN/ISSBN: 0090-5364
• DOI: 10.1214/21-AOS2095
• 发表日期: 2021
• 页码: 3510-3533
• 关键词: Nonparametric regression; riemannian framework; MANIFOLDS; geometry; models; matrices; images

摘要

Non-Euclidean data that are indexed with a scalar predictor such as time are increasingly encountered in data applications, while statistical methodology and theory for such random objects are not well developed yet. To address the need for new methodology in this area, we develop a total variation regularization technique for nonparametric Frechet regression, which refers to a regression setting where a response residing in a metric space is paired with a scalar predictor and the target is a conditional Frechet mean. Specifically, we seek to approximate an unknown metric-space valued function by an estimator that minimizes the Frechet version of least squares and at the same time has small total variation, appropriately defined for metric-space valued objects. We show that the resulting estimator is representable by a piece-wise constant function and establish the minimax convergence rate of the proposed estimator for metric data objects that reside in Hadamard spaces. We illustrate the numerical performance of the proposed method for both simulated and real data, including metric spaces of symmetric positive-definite matrices with the affine-invariant distance, of probability distributions on the real line with the Wasserstein distance, and of phylogenetic trees with the Billera- Holmes-Vogtmann metric.