Skip to content

Latest commit

 

History

History
22 lines (16 loc) · 1.09 KB

README.md

File metadata and controls

22 lines (16 loc) · 1.09 KB
Raw

PHP-2550-Project-2

This repository contains code and pdf output for project 2 of Practical Data Analysis, Fall 2023, Biostatistics Department, Brown University. Instructor: Alice Paul; Author: Yu Yan

  1. Objective: Develop regression models to predict tracheostomy placement or patient mortality in severe bronchopulmonary dysplasia (sBPD) cases. Data Source: BPD Collaborative Registry, including demographic, diagnostic, and respiratory parameters.

  2. Methodology:

    • Data assessment for quality and completeness.
    • Exploratory data analysis with transformations, variable selection, and outlier detection.
    • Construction of regression models.

  3. Models Presented: Lasso and Best Subset regression models.

  4. Performance: Both models show strong performance in various metrics.

  5. Limitations & Future Directions:

  • Explore additional machine learning methods (e.g., RandomForest).
  • Consider multilevel mixed-effect models for center-related effects.
  • Address the challenge of missing 44-week data.
  • Evaluate model sparsity vs. predictive power trade-offs for practical clinical use.