Vera Maya Santi, Aulia Desvira, Dania Siregar
Data on severe pneumonia among toddlers is binary data that only has two categories. One of regression methods that is often used in binary data is Binary Logistic regression. Regression modeling with many predictor variables causes multicollinearity problems in the data. Multicollinearity in the data causes the parameter estimates to be insignificant. In overcoming multicollinearity problems, the use of penalty functions can be a solution. This study applies the Least Absolute Shrinkage and Selection Operator (LASSO) penalty to estimate parameters and select significant variables in the Binary Logistic regression model. This study aims to analyze the factors that influence severe pneumonia in toddlers. The study utilizes the LASSO-penalized Binary Logistic regression model on data related to severe pneumonia among children under five years old in the urban villages of the public health center zone in Palu district during 2021. Comparing the AIC and BIC values, it was observed that the LASSO penalized Binary Logistic regression model outperformed the conventional Binary Logistic regression in addressing multicollinearity challenges. Notably, variables such as toddler malnutrition, stunting, underweight, exclusive breastfeeding, BCG immunization, measles immunization, access to drinking water, toddler health services, and health worker ratio significantly influenced the model. The LASSO-penalized Binary Logistic regression model exhibited a classification accuracy rate of 78.3% and an error rate of 21.7%. © 2026 Author(s).
Statistics Study Program, Faculty of Mathematics and Natural Sciences, Universitas Negeri Jakarta, Jl. Rawamangun Muka, Kota Jakarta Timur, DKI Jakarta, 13220, Indonesia
Research at a Glance
Register to unlock
Topics & SDG Alignment
Register to unlock
Collaboration
Register to unlock
Author Profile (Selected)
Register to unlock
References Overview
Register to unlock
Journal & Source
Register to unlock
Metadata & Integrity
Register to unlock