Developed an unsupervised deep learning
abdelkarim elmourad
Developed an unsupervised deep learning pipeline using the NASA IMS Bearing Dataset. Trained an autoencoder on healthy 20 kHz vibration signals and used reconstruction error, statistical time-domain features, a 99th-percentile threshold, and a stable alarm rule to detect bearing degradation.
Like this project
Posted Jun 8, 2026
Developed an unsupervised deep learning pipeline using the NASA IMS Bearing Dataset. Trained an autoencoder on healthy 20 kHz vibration signals and used rec...
Likes
0
Views
0
