Clinically Explainable Machine Learning Models for Complication Risk in Pediatric Orbital Cellulitis: A Study on Literature-Guided Synthetic Data

This repository contains the fully synthetic dataset used for the study titled “Clinically Explainable Machine Learning Models for Complication Risk in Pediatric Orbital Cellulitis: A Study on Literature-Guided Synthetic Data” The dataset consists of 2,000 artificial patient records generated using literature-guided statistical simulation. No real patient data were used and no identifiable information exists. Variables include demographics, clinical presentation, laboratory markers, imaging findings, and a composite complication outcome. The dataset is intended solely for methodological research, reproducibility, and validation of machine learning workflows.