Model Regresi Weibull Pada Data Kontinu yang Diklasifikasikan

Studi Kasus: Indikator Pencemaran Air BOD di DAS Mahakam Tahun 2016

Abstrak

Weibull Regression is a model of regression developed from Weibull distribution in which scale parameter is expressed in the regression parameters. The Weibull regression models discussed in this study are the Weibull survival regression, Weibull hazard regression and regression model for the mean. The Weibull survival regression model is a model of the probability that the Mahakam River water is polluted. The Weibull hazard regression model is a model of velocity of the polluted Mahakam River water, and the Weibull regression for the mean is the model used to predict the average value of BOD (Biochemical Oxygen Demand).  The purpose of this study was to obtain the Weibull regression model on BOD water pollution indicator data in the Mahakam River basin, to determine the factors that influence the Weibull regression model. The parameter method is maximum likelihood estimation (MLE). Based on the parameter estimation results, the maximum likelihood estimator is obtained by using the method of Newton-Raphson iteration. The results of hypothesis testing, it is concluded that the factors that influence the Weibull regression model are pH, Total Dissolved Solid (TDS) and water discharge.