ESTIMASI PARAMETER REGRESI LOGISTIK MULTINOMIAL MENGGUNAKAN METODE MAKSIMUM LIKELIHOOD

Logistic multinomial regression is a logistic regression analysis wich the dependent variable has a polychotomous or multinomialscale divided into two categories. The parameter estimation of the logistic multinomial regression model used Maximum Likelihood method and iteration Newton Raphson. The study use primary data from medical records of patients with a sample of 90 respondents. Tuberculosis inpatient patients at the Temanggug General Hospital are used as data for this study, consisting of three dependent variables namely Lung Tuberculosis (1), Bone Tuberculosis (2), Laryngitis Tuberculosis (3) and seven independent variables namely age