Rigorous Out-of-Sample Evaluation of a Global Phenomenological Model for Galaxy Rotation Curves Using Monte Carlo Cross-Validation

This repository contains all data, figures, and outputs associated with the manuscript: "Rigorous Out-of-Sample Evaluation of a Global Phenomenological Model for Galaxy Rotation Curves Using Monte Carlo Cross-Validation" The study presents a strict out-of-sample validation of a global six-parameter phenomenological model (UHST) using 100 Monte Carlo Cross-Validation (MCCV) iterations on the SPARC dataset. Contents: - Final manuscript (PDF)- MCCV summary table (train/test RMSE statistics)- Parameter stability table across MCCV runs- Visualization figures: - RMSE distribution (histogram) - Train vs test RMSE scatter plot The results demonstrate stable predictive performance (~23 km/s RMSE) with no evidence of overfitting, highlighting the importance of strict out-of-sample validation in galaxy rotation curve modeling. This repository ensures full reproducibility of all reported results.