adeleke1 / semiconductor-process-monitoring

A simple pass/fail yield classification for in-house testing based on collected signals/variable for a semiconductor manufacturing process

Semiconductor Manufacturing Process Monitoring

• The dataset and data information was taken from the UCI ML Repository. You will need all three files from the Data Folder link.
• Dataset was cleaned to handle missing data.
• Correlation coefficient to determine which features are correlated.
• Logistic Regression model was used to predict if there is a fail or pass.
• Important features were determined based on coefficient values and the model was rebuilt.
• Performance comparison of the two models using Confusion Matrix.