Claim Missing Document
Check
Articles

Found 2 Documents
Search
Journal : JITEK (Jurnal Ilmiah Teknosains)

ANALISIS SURVIVAL MODEL REGRESI SEMIPARAMETRIK PADA LAMA STUDI MAHASISWA Novita Eka Chandra; Siti Alfiatur Rohmaniah
Jurnal Ilmiah Teknosains Vol 5, No 2 (2019): JiTek
Publisher : Universitas PGRI Semarang

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (307.681 KB) | DOI: 10.26877/jitek.v5i2.4256

Abstract

In survival analysis to determine the relationship between variables is used a regression model, one of which uses the semiparametric regression model. The semiparametric regression model is a model that does not require assumptions or information on survival data distribution. That way, this model is more flexible in its use. In this study, the semiparametric regression model used the Cox Proportional Hazard (Cox PH) regression model. Estimation of Cox PH regression parameters can be done without determining the function baseline hazard. The purpose of this study is to determine the factors that influence the duration of student studies. If there are many students whose studies have not been on time, it shows that there is a lack of professionalism in the academic field of the educator. Thus, the community will assess the low quality of the university, resulting in a decrease in the number of students who want to study at the university. The samples in this study were students of class 2014 Universitas Islam Darul Ulum Lamongan. The variables have used the length of study for students, gender, GPA, school origin, organization, and work. Based on the results of the assumption Proportional Hazard (PH) conducted, all independent variables have fulfilled these assumptions, so that these variables can be used in Cox PH regression. After parameter estimation by Cox PH regression, the GPA and organizational factors significantly influence the duration of student study. Students with high GPA and participating in organizations more quickly complete their studies.
PERHITUNGAN PREMI ASURANSI JIWA MENGGUNAKAN GENERALIZED LINEAR MIXED MODELS Siti Alfiatur Rohmaniah; Novita Eka Chandra
Jurnal Ilmiah Teknosains Vol 4, No 2 (2018): JiTek
Publisher : Universitas PGRI Semarang

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (321.266 KB) | DOI: 10.26877/jitek.v4i2.3004

Abstract

The price of life insurance premiums for each person depends on the probability of death, not only based on age and gender as offered by an Indonesian insurance company.  The purpose of this study is to determine premium prices on underwriting factors and frailty factors using Generalized Linear Mixed Models (GLMM). GLMM is used for modeling a combination of fixed effect heterogeneity (underwriting factors) and random effects (frailty factors) between individuals. The data used longitudinal data about underwriting factors that have Binomial distribution are taken from the Health and Retirement Study and processed using R software. Because the data used by survey data within an interval of two years, so the probability of death is obtained for an interval the next two years. Underwriting factors that have a significant effect on the probability of death are age, alcoholic status, heart disease, and diabetes. As a result, is obtained the probability of death models each individual to determine life insurance premium prices. The premium price of each individual is different because depends on underwriting factors and frailty. If frailty is positive, it means that a person level of vulnerability when experiencing the risk of death is greater than negative frailty.