Water Quality Predictions Based on Grey Relation Analysis Enhanced LSTM Algorithms
, Zhenlin Wang 1, Elias Taalab 1, Baofeng Zhang 3, Xiaodong Li 4, Jiyong Wang 5, Muk Chen Ong 6 and Zefei Zhu 1
Round 1
Reviewer 1 Report
Line 30: because the study area is Qingdao lake, it is better to change “cities” into “basins “
Line 50 & 54： “Error! Reference source not found..”， Please check the manuscript before submission.
In the introduction, it is better to tell the story from the perspective of ensuring drinking water safety, rather than from the water pollution control. In the last paragraph, the Grey Relation Analysis (GRA) is introduced after your research aim, it is better to describe the advantages and disadvantages of GRA based on references review as you did for LSTM.
The basic information (size, statistics) of training set, developing set and validation set should be clearly described, besides figure 4.
Normally, when comparing different algorithms, the computing cost should also be mentioned in the same hardware and software configurations.
Figure 2: The information is too limited for international readers. Please enrich it with necessary information. It is easy to find good examples in international journals.
The quantitative results of RSME of LSTM and GRA-LSTM should be described and discussed in the Result section or a new section of DISCUSSION which discuss the advantages of limitations of the proposed GRA-LSTM, and the academic gaps you have filled.
In addition to root mean square errors, other widely used indicators such as Mean absolute error, Nash coefficient etc. in model comparison are also recommended to include in revised version.
Author Response
We thank the Reviewer for the detailed suggestions. We carefully considered all the questions raised by the Reviewer and answered them point by point in the following. Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The paper studies the Water Quality Predictions Based on Grey Relation Analysis En-2 hanced LSTM Algorithms. Overall, the paper is written well, however, a few comments should be considered before accepting the paper for publication.
1. Introduction needs to be rewritten again. The authors need to strength their arguments regarding the importance of their study by summarizing the previous studies focusing on the same research idea.
2. Lines 50 and 54 Check the reference. Errors.
3. Lines 65-74: The novelty needs to be highlighted better. What is the new offered for readers in this study?
4. Figure 2 is not important in my opinion. It can be deleted.
5. Qulaity of Figure 5 is not good.
Author Response
We thank the Reviewer for the detailed suggestions. We carefully considered all the questions raised by the Reviewer and answered them point by point in the following. Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
The manuscript concerns the important issue of the water quality predictions based on grey relation analysis enhanced LSTM algorithms. In the work, 2190 sets of monitoring data from automatic water quality monitoring stations in the Qiandao Lake, China from 2019 to 2020 were collected and served as training samples for prediction models. A grey relation analysis enhanced- long short-term memory (GRA-LSTM) algorithm is used to predict the key parameters of drink water quality. Remarks: Better quality of the Figure 2. Geographical position of the water quality monitoring station should be provided. Line 50, 54: Error! Reference source not found.)’’ these messages should be replaced by the the correct references. Instead of asterix * in the equations, use “x” or “∙”. Please, explain the value added of the mansucript with respect to existing methods. Add some perspectives of the future work concerning the modelling the system behaviour when applying the Markov chain with a discrete parameter, and the other one applies a state space model, as for example: Modelling water distribution network failures and deterioration, 2017, IEEE International Conference on Industrial Engineering and Engineering Management 2017-December, pp. 924-928. Are there concrete steps that can be recommended after performed research? If possible, please add in the conclusions a list of the main findings and interpretation, practical recommendations, and detailed research perspectives.
Author Response
We thank the Reviewer for the detailed suggestions. We carefully considered all the questions raised by the Reviewer and answered them point by point in the following. Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
All my comments and suggestions in the first-round review are carefully responded. I do not any serious comments and suggestions except for :
(1) The units of water quality parameters are missing in table 1;
(2) The fonts are too big in figure 2.
(3) Figure 3 should be improved, in the line to performance evaluation, there should be a word “yes”. And “No” and “yes” should not cut the lines. The rectangles surrounding “Training set”, “Developing set”, and “Validation set” should be central aligned.
(4) The titles of table 1 and table 2 are the same! I think, the title of table 2 is wrong.
(5) It is better not to include any references in conclusion section.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
Accept in present form.
