A Lightweight Bearing Fault Diagnosis Method Based on Multi-Channel Depthwise Separable Convolutional Neural Network
Round 1
Reviewer 1 Report
The paper is interesting and free of major mistakes.
There are some issues that should be improved:
1. Several tables (like 8, 9, 10, 14 …) provide performance results but do not tell which of the performance metric is presented. This should be supplemented. Otherwise, those results are unreadable.
2. The manuscript should be supplemented with a description of the evaluation process, with an emphasis on the metric calculation – what metrics are used and how they are calculated for this specific research, and how those results will be interpreted.
3. Table 13 presents the accuracy of classifiers for different subsets. One thing is the accuracy metric is not the best way to treat such an evaluation, it is not robust, and there are better metrics, like F1-score or MCC, that provide a resilient image of the evaluated model. Such a method evaluation using accuracy metric only is not sufficient. Please replace accuracy results with F1 or MCC metric, or at least append them to performance tables.
4. Section 4.1 provides no information on dataset availability. Similarly, for the code of the models, are they available publicly or for demand?
And some minor issues:
5. Some unwanted characters in the title – should be removed
6. Section 2 presents GELU function but does not fully define it, i.e., what G stands for? Where the factor 0.044715 is coming from?
7. For table 1 – the label “/128” is unclear, shouldn’t be “- / 128” ?
Best wishes
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
The authors proposed a lightweight multi-channel deep separable convolutional neural network (MCDS-CNN) for rolling bearing fault diagnosis.
The work is interesting but needs some editing for improving the English language. Could you add a comparison with other state-of-the-art methods for rolling bearing fault diagnosis on the same dataset?
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
DD-CNN stands for what? Please add the reference (if any).
Author Response
Please see the attachment.
Author Response File: Author Response.pdf