Physics-Informed Stacking Ensemble for Simultaneous Prediction of Surface Roughness and Sound Pressure Level in VBD Marble Milling — Dataset and Code

Experimental datasets (2,640 observations each for Afyon White and Afyon Grey marble), Google Colab notebooks, and all reproducible output files for the physics-informed stacking ensemble study. The pipeline combines Boruta-SHAP feature selection, chronological blocked cross-validation, and a Ridge meta-learner to simultaneously predict surface roughness (Ra) and sound pressure level (Lp) in VBD marble milling.