一种基于核函数的函数型数据非参数回归方法  

A Kernel Functions Based Nonparametric Functional Data Regression Approach

作　　者：柳心阳 李秀英[1] 耿发展[1] LIU Xinyang;LI Xiuying;GENG Fazhan(School of Mathematics and Statistics,Changshu Institute of Technology,Changshu 215500,China)

机构地区：[1]常熟理工学院数学与统计学院,江苏常熟215500

出　　处：《常熟理工学院学报》2025年第2期103-106,共4页Journal of Changshu Institute of Technology

摘　　要：函数型数据分析因其在不同领域的广泛应用而受到统计学习的广泛关注,现有的函数型数据回归方法大多集中在线性模型上,非线性函数型数据回归的相关研究较少.本文基于再生核函数提出一种新的函数型数据非参数回归方法,并通过数值实验验证了所提出的方法的有效性和鲁棒性.Functional data analysis has attracted much attention in statistics learning due to its wide applications in different fields.Most of existing functional data regression approaches focus on linear models.And the discussion on nonlinear functional data regression is rare.In this paper,our goal is to propose a novel nonlinear regression approach for functional data analysis based on reproducing kernel functions.Some numerical experiments are performed to show the effectiveness of the proposed approach.

关 键 词：函数型数据分析 非线性回归方法 核方法 非参数回归 

分 类 号：O175[理学—数学]