Uncertain Semi-parametric Additive Model Based on Orthogonal Series Method

In engineering applications, uncertain nonparametric and semi-parametric regression models are of interest for effectively exploring complex correlations of imprecise data in small samples. This paper proposes a new model: the uncertain semi-parametric additive model, which incorporates multidimensi...

Full description

Saved in:
Bibliographic Details
Published in2024 IEEE 25th China Conference on System Simulation Technology and its Application (CCSSTA) pp. 341 - 345
Main Authors Aizitili, Aikebaier, Zhang, Huiguo
Format Conference Proceeding
LanguageEnglish
Published IEEE 21.07.2024
Subjects
Additives
Analytical models
Data models
Estimation
imprecise observations
least squares estimation
Numerical models
Numerical simulation
orthogonal series method
Predictive models
Regression analysis
Solid modeling
Systems simulation
uncertain semi-parametric additive model
Online AccessGet full text
DOI10.1109/CCSSTA62096.2024.10691726

Cover

More Information
Summary:In engineering applications, uncertain nonparametric and semi-parametric regression models are of interest for effectively exploring complex correlations of imprecise data in small samples. This paper proposes a new model: the uncertain semi-parametric additive model, which incorporates multidimensional explanatory variables into the analytical process and sets both linear and nonlinear regression relationships. The parametric and nonparametric regression functions of the new model are fitted using uncertain least squares and orthogonal series methods, respectively. This paper also investigates the residual analysis, forecast value and forecast interval of the uncertain semi-parametric additive model. Finally, the validity and applicability of the proposed model and method are verified by numerical simulation of a series of small-sample observations in practical engineering applications.
DOI:10.1109/CCSSTA62096.2024.10691726