EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

A numerical comparison between lifetime regression models and finite normal mixture regression model

P.V. Sivadas

Malaya Journal of Matematik · 01.10.2019 · 7 (04) · 892-897

DOI: 10.26637/mjm0704/0044 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Malaya Journal of Matematik, online erschienen am 01.10.2019. Verfasst von P.V. Sivadas. Die Arbeit nennt 24 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This study examines age at stroke onset using probabilistic and regression-based models to capture population heterogeneity in Estonian stroke data (Korvet et al., 1997). Finite normal mixture models are fitted using the Expectation-Maximization (EM) algorithm and compared with Weibull, Gamma, Lognormal, and Log-logistic distributions. Based on the Akaike and Bayesian information criteria, a three-component normal mixture model provides the best fit, indicating the presence of latent subgroups in the stroke population. To assess age-related stroke risk, a Weibull accelerated failure time (AFT) regression model is fitted to the exponentiated age data, and a normal mixture regression model with linear predictors is estimated using the EM algorithm to account for unobserved heterogeneity. The results demonstrate the effectiveness of mixture-based and regression models in capturing complex patterns in stroke age data.

Zitieren

APA

Sivadas, P. (2019). A numerical comparison between lifetime regression models and finite normal mixture regression model. Malaya Journal of Matematik, 7(04), 892-897. https://doi.org/10.26637/mjm0704/0044

Vancouver

Sivadas P. A numerical comparison between lifetime regression models and finite normal mixture regression model. Malaya Journal of Matematik. 2019;7(04):892-897. doi:10.26637/mjm0704/0044

BibTeX
@article{Sivadas2019,
  author = {Sivadas, P.V.},
  title = {A numerical comparison between lifetime regression models and finite normal mixture regression model},
  journal = {Malaya Journal of Matematik},
  year = {2019},
  volume = {7},
  number = {04},
  pages = {892--897},
  doi = {10.26637/mjm0704/0044}
}
Kostenlos starten →