Prediction and Optimization of Surface Roughness in Electrical Discharge Machining Using Machine Learning Models

Abstract:Electrical Discharge Machining (EDM) is a widely used non-traditional machining process for shaping and finishing hard-to-machine materials. Achieving an optimal surface roughness is challenging due to the nonlinear influence of multiple EDM parameters. This research applies machine learning techniques to predict and optimize surface roughness using key process variables such as current, pulse-on time, pulse-off time, and voltage. Various regression models—including Linear Regression, Decision Tree Regression, Random Forest Regression, and Support Vector Regression (SVR)—were analyzed for performance. Results show that SVR achieved the highest accuracy with the lowest mean absolute error and highest R² score. Hyperparameter tuning further improved model performance. The study demonstrates that machine learning provides a reliable and effective method for optimizing EDM parameters, contributing to improved machining precision and product quality.