Research on Intelligent Grading Evaluation of Water Conservancy Project Safety Risks Based on Deep Learning

Round 1

Reviewer 1 Report (Previous Reviewer 2)

Accept as is

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report (New Reviewer)

Grading evaluation of water conservancy project safety risks was studied in this study. It helps to address the practical problem in the water conservancy project. From the methodological side, this paper integrates Priori Attention and constructs Transformer risk prediction model based on a sliding window. The studied topic is important and the proposed method is interesting. The reviewer has the following comments.

1. The literature review should be improved with more recently published works reviewed.

2. The research gap should be well described. Thus, the main limitations of current research should be discussed.

3. The benchmarking methods should include newly published methods.

4. The hyper-parameters of the deep learning model should be optimized.

5. Some typos are observed and the figure quality should be improved.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report (New Reviewer)

The paper has been well improved and is ready for publicaiton. 

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

After reading the manuscript, I had the impression, in fact I am convinced, that the authors had directed their article to the wrong journal. I believe that the topics described in the work should be presented for publication in periodicals concerning the development of machine learning algorithms, risk analysis in industrial processes or project management processes, but certainly not to a journal dealing with water resources processes, such as Water (Section Hydraulics and hydrodynamics) is.

Therefore, I suggest the publisher to reject this article at the pre-decision stage and to suggest to the authors another journal (possibly from MDPI) to which they should submit their work.

In addition to this general conclusion, I would like to point out that the article itself has, in my opinion, some shortcomings that would be worth correcting before resubmitting the article. These are, for example:

- In the literature review, especially when describing the current state of knowledge (e.g. lines 103-120), one can see the dominance of works from China and the surrounding area, which would suggest that in other parts of the world, there is no significant research on the subject. A more complete review of the literature is desirable.

- In the Literature Review section (lines 67-94), the authors use strange ways of citing literature sources, such as Literature [No] provides, or Scholar Name [No], rather unheard of in scholarly articles.

- The description of the Study area in the Materials and Methods section is poor. We do not learn much here about the details of hydrotechnical projects, there is not even a map with the location of the considered investments. This also indicates that in this article the issues of water engineering are in the background of the authors' interest.

At the same time, I would like to point out that I do not assess the correctness of the proposed machine learning algorithm.

Reviewer 2 Report

The paper needs major revisions:

1-Literature review on "n Intelligent Grading Evaluation of Water Conservancy Project Safety Risks Based on Deep Learning" requires more citations in the introduction section. Additionally, a brief conclusion of literature review is essential.

2-Properties of the understudy flood events can be more extended in the study area. 

3-In line 248: "is" would read "are". the paper need major refinements in term of English

4-How were weighting coefficients computed in Eqs.(5) to (7)??

5-Setting parameters of In the model prediction correctness experiments, multiple network models (SVM, 299 CNN, GAT, RCNN, Transformer) should be detailed and also general descriptions of Machine Learning models should be included.