PDA-project2

The project2 completed by Liangkang Wang

Neonatal Care Prediction Models

This repository hosts the R Markdown and related analysis files for a project focused on predicting critical outcomes in neonatal care. Utilizing Lasso and Stepwise logistic regression models, the project evaluates and compares the predictive accuracy and calibration of these models based on a clinical dataset.

Overview

The project includes:

• Implementation of Lasso and Stepwise logistic regression models.
• Evaluation of model performance using ROC-AUC, Brier Scores, and Hosmer-Lemeshow tests.
• Variable importance analysis with detailed model interpretation.
• Data handling using Multiple Imputation by Chained Equations (MICE).