An Expandable Yield Prediction Framework Using Explainable Artificial Intelligence for Semiconductor Manufacturing

Enormous amounts of data are generated and analyzed in the latest semiconductor industry. Established yield prediction studies have dealt with one type or a dataset from procedure. However, device fabrication comprises hundreds processes, various factors affect yields. This challenge is addressed this study by using an expandable input data-based framework to include divergent adapting explainable artificial intelligence (XAI), which utilizes model interpretation modify conditions. After preprocessing data, procedure optimizing comparing several machine learning models followed select best performing for dataset, random forest (RF) regression root mean square error (RMSE) value 0.648. The results enhance production management, explanations deepen understanding yield-related Shapley additive explanation (SHAP) values. work provides evidence empirical case data. improves accuracy, relationships between features illustrated SHAP value. proposed approach can potentially analyze fields conditions interpret multifaceted manufacturing.