Skip to content

Latest commit

 

History

History
30 lines (18 loc) · 535 Bytes

File metadata and controls

30 lines (18 loc) · 535 Bytes

Machines_Predict_PCA

Applying PCA and Logistic Regression to data industrial machines to provide maintenance.

Dataset: ficticius data sensors machine - 1000 rows with target.

alt text

Pre-processing:

  • StandScaler

Find Most relevant variables:

  • Principal component analysis (PCA)

Model Prediction:

  • Logistic Regression

Cross Validation

  • LeaveOneOut (LOOCV)

Results

alt text alt text

alt text

Simulation:

alt text