The diagnosis/prognosis problem has already been introduced by the authors in previous papers as a classification problem for survival data. In this paper, the specific aspects of the estimation of the survival functions in diagnostic classes and the evaluation of the posterior probabilities of the diagnostic classes are addressed; a latent random variable Z is defined to denote the classification of censored and uncensored individuals, where early censored individuals cannot be immediately classified as Z is not observed. Parameter estimation of the mixture survival model thus derived is carried out using a proper version of the EM algorithm with given prior probabilities on Z and diagnostic/prognostic information provided by the observable covariates is also included into the model. Numerical examples using AIDS data and a simulation study are used to better outline the main features of the model and of the estimation methodology.

Schinaia, G., Rossi, C. (2000). EM estimation of diagnosis. BIOMETRICAL JOURNAL, 42(5), 583-604.

EM estimation of diagnosis

ROSSI, CARLA
2000

Abstract

The diagnosis/prognosis problem has already been introduced by the authors in previous papers as a classification problem for survival data. In this paper, the specific aspects of the estimation of the survival functions in diagnostic classes and the evaluation of the posterior probabilities of the diagnostic classes are addressed; a latent random variable Z is defined to denote the classification of censored and uncensored individuals, where early censored individuals cannot be immediately classified as Z is not observed. Parameter estimation of the mixture survival model thus derived is carried out using a proper version of the EM algorithm with given prior probabilities on Z and diagnostic/prognostic information provided by the observable covariates is also included into the model. Numerical examples using AIDS data and a simulation study are used to better outline the main features of the model and of the estimation methodology.
Pubblicato
Rilevanza internazionale
Articolo
Sì, ma tipo non specificato
Settore MED/01 - Statistica Medica
English
Senza Impact Factor ISI
Diagnosis; EM algorithm; Latent variables; Mixture models; Survival analysis
Schinaia, G., Rossi, C. (2000). EM estimation of diagnosis. BIOMETRICAL JOURNAL, 42(5), 583-604.
Schinaia, G; Rossi, C
Articolo su rivista
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/2108/49645
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact