EEG Emotion Recognition Based on Federated Learning Framework

Round 1

Reviewer 1 Report

The article presents several formatting problems. Please add a space between the brackets of the citations [] and the words that precede them. For example, systems[1] should be systems [1]. Check throughout the document for the same errors. Figure 1 is presented twice and is not mentioned in the text. The equations are part of the text. Therefore, they should end with a comma or period. There are reference errors on lines 122, 180 and 272. No related work of the last year is referenced. It is recommended to read the following paper:

- Li, X., Zhang, Y., Tiwari, P., Song, D., Hu, B., Yang, M., ... & Marttinen, P. (2022). EEG based Emotion Recognition: A Tutorial and Review. ACM Computing Surveys (CSUR).

 

 

 

Author Response

Dear Reviewers, Thank you for your comments on our article. We have made changes to the article based on your comments. We have uploaded the "Respond Reviewer1" file in PDF format, where we have answered your questions in detail, explained how the article was revised and added the revised article to the back of the file. Please see attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

1.     Line 121: reference error.

2.     Line 180: reference error.

3.     Line 272: reference error.

4.     Line 225: repetitive “receiving”

5.     Accuracy is used as the only metric in the experiments. Since this is multi-classification problem, why precision, recall and f1 score are not calculated for comprehensive assessment?

6.     Line 264-265: please clarify the which threshold the authors refer to and why the values are different.  

7.     Why are there some 0 values in table 3? Will they affect the results?

8.     Line 208: What are the training parameters involved here? How are they determined?

 

9.     Line 280-286: Why the data are divided into these four datasets? Why not divide the data in an increasing quantity manner to validate the influence of different amount of data on prediction performance, for example, 25%, 50%, 75% and 100% of all subjects? Please clarify the rationale behind this setting.

Author Response

Dear Reviewers, Thank you for your comments on our article. We have made changes to the article based on your comments. We have uploaded the "Respond Reviewer2" file in PDF format, where we have answered your questions in detail, explained how the article was revised and added the revised article to the back of the file.Please see the attachment.

Author Response File: Author Response.pdf