Arcing-Issue-Solution-with-Random-Forest

Acring Issue Data Analysis Solution for Glass Panel Industry

1. Business Understanding: Analyzed glass panel FDC data with Python, conducted data mining flow
2. Data Mining and Model Tuning: Predicted the arcing classification and identified key factors with Random Forest model, optimized the model by Grid Search and Cross Validation
3. Model Evaluation: Plotted the Confusion Matrix and ROC Curve, achieved the accuracy of 0.875 and AUC of 0.92
4. Key Factors Identification: Identified the most important key factors and completed visualization according to variance importances
5. Demo Reporting: Solved the problem of dealing with wide data and multicollinearity, provided data analysis solution and gave a presentation to show the demo