MODELING OF SPATIAL BINARY LOGISTIC REGRESSION ON EMPLOYEE HEALTH DATA

• Main Author: Ahsanul Karimah, Shofie
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/49565
• Institution: Institut Teknologi Bandung

Cholesterol is a fat compound that is produced naturally by the liver but can also be found in foods of animal origin, such as meat and milk. While total cholesterol is a combination of the amount of good cholesterol, bad cholesterol, and triglycerides in units of milligrams per decilitre of blood. High cholesterol can cause clots of cholesterol in blood vessels, which can lead to heart attacks and strokes. Therefore, we want to know how influential factors that cause high cholesterol in employees' cholesterol conditions by using spatial logistic regression methods. Response variable is the total cholesterol condition of employees with normal and high categories. Predictor variables or factors suspected to cause high cholesterol are age, sex, BMI, triglycerides, fat, visceral fat, years of service, and spatial variables. The spatial factor uses the element of employee closeness in a site. Based on the analysis it is known that 41 percent of employees have high total cholesterol. In binary logistic regression modelling, the most influential variables causing high total cholesterol are triglyceride levels, spatial variables, fat content, age 51-65 years, and BMI in obesity category with a model accuracy rate of 70.8%.