Determining confidence interval and asymptotic distribution for parameters of multiresponse semiparametric regression model using smoothing spline estimator

Budi Lestari, Nur Chamidah, I. Nyoman Budiantara, Dursun Aydin

Research output: Contribution to journalArticlepeer-review

3 Citations (Scopus)

Abstract

The multiresponse semiparametric regression (MSR) model is a regression model with more than two response variables that are mutually correlated, and its regression function is composed of parametric and nonparametric components. The study objectives are propose a new method for estimating the MSR model using smoothing spline. Also, find the confidence interval (CI) of parameters and the distribution asymptotically of the model parameters estimator. Methods used in this study are reproducing kernel Hilbert space (RKHS) method and a developed penalized weighted least squares (PWLS), and apply pivotal quantity, central limit theorem, and theorems of Cramer-Wold and Slutsky. The results are an 100(1–α)% CI estimate and an asymptotic normal distribution for the parameters of the MSR model. In conclusion, the estimated MSR model is a combined components estimate of parametric and nonparametric which is linear to observation, and CIs of parameters depend on t distribution and estimator of parameters is asymptotically normally distributed. Future time, this study results can be used as theoretical bases to design standard growth charts of the toddlers which can then be used to assess the nutritional status of the toddlers.

Original languageEnglish
Article number102664
JournalJournal of King Saud University - Science
Volume35
Issue number5
DOIs
Publication statusPublished - Jul 2023

Keywords

  • Asymptotic distribution
  • Confidence interval
  • Nutritional status
  • Semiparametric regression
  • Smoothing spline

Fingerprint

Dive into the research topics of 'Determining confidence interval and asymptotic distribution for parameters of multiresponse semiparametric regression model using smoothing spline estimator'. Together they form a unique fingerprint.

Cite this