Machine Learning for Predicting Porosity and Circularity in Laser Powder Bed Fused Ti-6Al-4V

In this study, three machine learning (ML) models—backpropagation neural network (BP), random forest (RF), and support vector machine (SVM)-were employed to predict the porosity and circularity of laser powder bed fusion (LPBF) processed Ti-6Al-4V (TC4) samples under 75 different parameters. To evaluate the predictive performance of models, R square (R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>), mean absolute error (MAE), root mean squared error (RMSE) were calculated. Among the models, the SVM model demonstrated great accuracy in predicting both porosity and circularity. After further optimization of the SVM model using grid search and k-fold cross validation, the model achieved R square (R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>) of 0.945 for porosity and 0.819 for circularity in validation experiments, indicating great predictive capabilities.