Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Please consider the following comments while revision of this article.

 

1. Why IPDF and PSD are chosen for the preliminary classification stage?

2. Can you provide more details on the types of acceleration data anomalies that the preliminary CNN is designed to identify and classify?

 

3. How is the performance of the preliminary classification stage evaluated, and what are the key performance metrics? Please indicate explicitly.

4. What is the RPV indicator, and how is it calculated from the normal and outlier categories?

 

 

5. How are threshold values determined for the RPV indicator to achieve precise classification?

6. Are there any specific challenges or considerations in setting these threshold values?

7.  Can you explain the decision-making process for classifying data points as normal or anomalous based on the two-stage model?

8. How well does the model generalize to different datasets and scenarios beyond the training data?

 

 

 

Comments on the Quality of English Language

English language can be improved. 

Author Response

See Attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The paper summarizes the previous research and the proposed classification model effectively. In general, the authors have presented their approach in a clear way and overall with only minor issues:

1.       A general spelling and grammar review of the paper is suggested for minor corrections.

2.       Section 2.2.2.1, last paragraph explains the IPDF features of different data patterns shown in Figure 3. However, the biased class is missed. Please explain that data pattern as well.

3.       Figures 4 and 6 are not referenced in the text. Please incorporate those figures and their explanations into the corresponding text.

4.       Section 3.2 (line 265): “As shown in Equations (2) ~ (5), precision, recall, and F1 score are …” Accuracy should be included in that sentence as well.

5.       Figure 7: Legend shows “Train Accuracy” instead of Training Accuracy.

Comments on the Quality of English Language

Minor corrections

Author Response

See Attachment

Author Response File: Author Response.pdf