Calibrated Adaptive Teacher for Domain-Adaptive Intelligent Fault Diagnosis

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

  In this paper, the Calibrated Adaptive Teacher (CAT) method is proposed to improve the quality of pseudo-labels and reduce error accumulation. the predictions of the teacher network on target samples throughout the self-training process is calibrated by leveraging well-known post-hoc calibration techniques. The paper demonstrates a reasonable approach and effective methods. Minor revisions are suggested for further clarification.

1. Please explain the reasons why each transfer task has different difficulty levels from the perspective of signal characteristics based on the selected experimental data. Suggest analyzing the impact of the characteristics of different transfer tasks on the effectiveness of the proposed method.

 

2. Why is CAT + post-hoc calibration (such as MatrixScaling) not as effective as DANN or AT in some transfer tasks? The positive effect of the post-hoc calibration has not been reflected. Could you provide more analysis of the reasons?

Author Response

Dear reviewer,

Thank you for your feedback. Please find our detailed response in the attached PDF document.

Sincerely

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Please refer to attached document for reviews

Comments for author File: Comments.pdf

Author Response

Dear reviewer,

Thank you for your feedback. Please find our detailed response in the attached PDF document.

Sincerely

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Please find attached file of my reviews.

Comments for author File: Comments.pdf

Author Response

Please find the response to the review for round 2 in the attached document.

Sincerely

Author Response File: Author Response.pdf