Machine Learning Based Corrosion Prediction of Biodegradable Magnesium Alloys for Biomedical Implants

This study presents a data-driven machine learning framework for corrosion prediction of biodegradable magnesium alloys for biomedical implant applications. A dataset of 550 experimental observations containing 26 features was compiled from literature sources. Five machine learning algorithms — Random Forest, XGBoost, Gradient Boosting, Support Vector Regression, and Linear Regression — were implemented and evaluated. Random Forest achieved the highest prediction performance with an R² score of 0.90. Feature importance analysis revealed that processing condition, Zn content, Ca content, pH, and temperature are dominant factors influencing corrosion behaviour. The results demonstrate that machine learning models effectively predict corrosion resistance