Optimal Statistical Feature Subset Selection for Bearing Fault Detection and Severity Estimation

<div>Plots of the best IMFs selected on the basis of Pearson correlation coefficient. A drive-end sample of size 1024 points from each of the 12 classes has been depicted.</div>

Shock and Vibration

fig3

Figure 3

Figure 3: Optimal Statistical Feature Subset Selection for Bearing Fault Detection and Severity Estimation 