MODEL REGRESI COX PROPORSIONAL HAZARD PADA DATA DURASI PROSES KELAHIRAN DENGAN TIES

Triastuti Wuryandari(1*), Danardono Danardono(2), Gunardi Gunardi(3)


(1) 
(2) 
(3) 
(*) Corresponding Author

Abstract


Survival data are usually found in the fields of health, insurance, epidemiology, demography, etc. Survival data is characterized by a response in the form of time, one example is the duration of the birth process. The duration of the birth process is thought to be influenced by several factors, including the baby's weight, baby's height, mother's age, gestational age, gender and the method used to birth process. One of the regression models for survival data is the Cox regression proportional hazard model. Parameter estimation in the Cox regression is based on partial likelihood. If two or more individuals have the same survival value, it is called ties. If there are ties, then the partial likelihood will have problems in determining the risk set, so it is necessary to modify the partial likelihood. Methods that can be used to overcome ties are the Breslow, Efron and Exact methods. This method is a modification of parameter estimation using maximum partial likelihood. Parameter estimation results are obtained by maximizing the partial likelihood function using Newton Raphson iteration. The case study in this paper is data on the duration of the birth process. The best model for the duration of the birth process with ties is the Exact method because it has the smallest AIC value

Full Text:

PDF

Article Metrics

Abstract view : 75 times
PDF - 9 times

DOI: https://doi.org/10.26714/jsunimus.9.1.2021.47-55

Refbacks

  • There are currently no refbacks.


Copyright (c) 2021 Jurnal Statistika Universitas Muhammadiyah Semarang

Editorial Office:
Department of Statistics
Faculty Of Mathematics And Natural Sciences 
Universitas Muhammadiyah Semarang

Jl. Kedungmundu No. 18 Semarang Indonesia

Published by: 
Department of Statistics Universitas Muhammadiyah Semarang

View My Stats

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.