Yield Strength Prediction Using Machine Learning

Machine learning workflow for predicting yield strength using Linear, SVR, Random Forest and Gradient Boosting models in MATLAB.

This project provides a clear and reproducible MATLAB workflow for predicting yield strength (Sy) of engineering materials using machine learning techniques.

The script performs:

- Data cleaning and preprocessing

- One-hot encoding of categorical variables

- 80/20 holdout validation

- Model training and comparison

Implemented models:

• Linear Regression (Ridge)

• Support Vector Regression (Bayesian optimization)

• Random Forest (TreeBagger)

• Gradient Boosting (LSBoost)

Performance is evaluated using MAE, RMSE, and R² metrics.

A simple ablation experiment is also included to analyze feature influence.

This work was developed as part of a machine learning training program. The code is intentionally written as a readable baseline implementation suitable for educational use and model comparison.