We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Statistics > Machine Learning

Title: Quick Learning Mechanism with Cross-Domain Adaptation for Intelligent Fault Diagnosis

Abstract: The fault diagnostic model trained for a laboratory case machine fails to perform well on the industrial machines running under variable operating conditions. For every new operating condition of such machines, a new diagnostic model has to be trained which is a time-consuming and uneconomical process. Therefore, we propose a quick learning mechanism that can transform the existing diagnostic model into a new model suitable for industrial machines operating in different conditions. The proposed method uses the Net2Net transformation followed by a fine-tuning to cancel/minimize the maximum mean discrepancy between the new data and the previous one. The fine-tuning of the model requires a very less amount of labelled target samples and very few iterations of training. Therefore, the proposed method is capable of learning the new target data pattern quickly. The effectiveness of the proposed fault diagnosis method has been demonstrated on the Case Western Reserve University dataset, Intelligent Maintenance Systems bearing dataset, and Paderborn university dataset under the wide variations of the operating conditions. It has been validated that the diagnostic model trained on artificially damaged fault datasets can be used to quickly train another model for a real damage dataset.
Comments: 9 pages, 6 figures, transaction
Subjects: Machine Learning (stat.ML); Machine Learning (cs.LG)
Journal reference: IEEE Transactions on Artificial Intelligence, Nov, 2021
DOI: 10.1109/TAI.2021.3123935
Report number: TAI-2021-Mar-A-00107
Cite as: arXiv:2103.08889 [stat.ML]
  (or arXiv:2103.08889v2 [stat.ML] for this version)

Submission history

From: Arun Sharma Mr. [view email]
[v1] Tue, 16 Mar 2021 07:24:37 GMT (1163kb)
[v2] Mon, 6 Sep 2021 22:07:59 GMT (845kb,D)

Link back to: arXiv, form interface, contact.