bda-final

RUL Prediction with Random Forest Regression on CMAPSS FD004 dataset

• Piecewise RUL degeneration function with early RUL = 125
• Classification of operating conditions
• K-Means for fault modes target of the train set and SVM for fault modes classification of the test set
• Random forest regression for RUL prediction
• 2 baseline models:
  • RF regression with raw sensors' signal
  • Decision tree regression with the same feature engineering

Fault Classification with SVM on CWRU dataset

• Wavelet packet decomposition for signal processing and energy for featuring engineering
• SVM for classification