Analysis of the Composition of Ancient Glass and Its Identification Based on the Daen-LR, ARIMA-LSTM and MLR Combined Process

Round 1

Reviewer 1 Report

Figure 4 explanation is too long.  This information can be included in the text and the figure can be cited.

 Figure 2 description could be more detailed.  In addition, this figure can be further detailed and colored so that the reader can clearly understand the article at first glance.

 The article is very well prepared, but in the last part of the introduction, the difference of the study from other studies in the literal type should be given clearly.

 The article whose doi is given below must be cited.  The narrative flow in this article can serve as an example for writing this article.

 https://doi.org/10.1109/ACCESS.2021.3088612

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

The paper focuses on a novel idea with limitations. The following points need to be addressed in the revision:

1. Many typos and grammatical mistakes in the article need to be corrected.
2. Motivation needs to be projected in the abstract showing why this study is important.
3. Use short names, like BP, by writing the complete pronunciation at its first appearance.
4. I could not find any reference or details related to Deep Learning (DL). Although sophisticated DL models are being employed, add a section related to your trials with deep learning. Add the following reference related to designing and constructing materials using artificial intelligence.

 

https://doi.org/10.3390/bios12121181

 

5. The paper needs more literature surveys with state-of-the-art methods apart from the application. You should add references to the following recent articles in the introduction section:

https://ieeexplore.ieee.org/abstract/document/10086464

https://doi.org/10.54097/hset.v39i.6859

https://doi.org/10.23977/autml.2022.030210

https://doi.org/10.54097/hset.v22i.3367

 

6. In the methodology explained for Daen-LR, add details related to regularization carried out to cope with the complexity of the model in terms of penalty.
7. In line no. 193, the same line is repeated where we started earlier. Similarly, check on line no. 207, 217, 219, 233, 239, etc. for error.
8. In section 2.2.1, phi is not defined. Check for all such errors in the revision. Explain all variables and parameters in the equations shown to explain the algorithm.
9. The logic for 105% in line no. 305 needs to be clarified. Clarify this a bit more and try to avoid quoting such proportions.
10. Machine-specific details, platforms, and computational environment are missing.
11. Sometimes you are writing Dean (line 328) instead of Daen. Correct it throughout the article.
12. How were the outliers removed (mentioned in Table 2)? Add details to the article.
13. Figure 4 has never been referred to and discussed in the article.
14. Axis is not mentioned in Figure 5. Add the axis labels and explain them with error-free sentences.
15. Comparison with SOTA techniques needs to be included.
16. Performance measures used need to be detailed separately before results and discussion.
17. The conclusion needs revision.
18. Limitations and future work recommendations need to be added before the conclusion.

 

 

It needs to be improved.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

1. The current abstract only describes the general purposes of the article. It should also include the article's main impact and significance.

2. The authors should demonstrate that all issues of the tools used are being considered or are negligible and provide a strong motivation for using their proposed techniques.

3. Please more clearing state the contribution of your paper in the last paragraph of the introduction

 

4. Would like to see the literature review enhanced and the results and discussion compared with other recent studies.

5. Would like to see the discussion section providing some remarks to further discuss the proposed method, for example, what are the main advantages and limitations in comparison with existing methods?

6. What is exactly the input and output parameters in your research?

7. The presentation of the paper is too weak.

8. In conclusion section, the authors need to clearly show the goals, data, methods, major findings, implications, drawbacks, and future research. Please carefully add the related descriptions on implications and drawbacks and improve the other mentioned parts of the conclusion.

 

 

 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 4 Report

 

Here are some identifiable issues with the manuscript.

Lines 50-51: “In 1992, Korean scientist Lee Chul, his chemometric pattern recognition method was…” is awkwardly phrased.

Line 61: “Nadine Schibille and his team …” - should probably be “Nadine Schibille and her team …”

 

Lines 85-86: “...the official website of the 2022 China St.dent Mathematical Modeling Competition [14], Many…” - replace “[14], Many..” with “[14], Many..”

 

Lines 86 – 90: “..Many of these glasses have weathered under the influence of thou-

sands of years of wind and rain as well as humid burial environment, and the internal

elements of the glass have exchanged with the external elements of the environment in

large quantities, resulting in significant changes in the proportional content of its internal

chemical composition, thus relying on a simple method of determining the type of glass” – this is a long, run-on sentence. Please rewrite using shorter sentences.

 

Line 114: “.. Oracle properties and adaptive classification effects...” - explain and/or provide a citation to indicate to the reader what these are and why they matter.

 

Equation (1): “Logic(P)” should be “Logit(P)”

 

Lines 139 – 141: “The first order derivatives of in the above equation and make them equal to 0, respectively, can be found for all the coefficients to be determined in the system of equations β .” - why is “making them equal to zero” needed? A brief explanation will be helpful.

 

Line 145: “...properties [19]; Second,” either change the semi-colon to a period (full-stop), or change “Second” to “second”.

 

Line 154: “ Double adaptive elastic net Logistic ” convert all words to lower-case.

 

Lines 207, 217, 219, 233, 297: “! Reference source not found.” - this error should be resolved.

 

Lines 241 – 252: The LSTM model’s description is inadequate, and its introduction is abrupt. Please provide a clearer transition from the previous paragraph. Explain why this model is appropriate, highlight the key components and define them, before using them in your explanation.

 

Line 260: “nonlinear predictio n” should be “nonlinear prediction”

 

Line 288: “We performed BP test and White test” – what are these and why are these appropriate?

 

Lines 291-292: “the variance inflation factor VIF was used to eliminate the influence factors with multicollinearity” – how?

 

Lines 313 – 318: Replace semi-colons with periods.

 

Line 363: “Through research calculations” – what are these?

 

Line 380: “Assuming that the residual is a white noise sequence,” -- why is this a reasonable assumption?

 

Line 530: Remove space between “follows” and the colon following it.

 

Logistic regression’s validity in the analysis of a particular dataset is predicated on the validity of the assumptions associated with the model. Apart from checking for the presence of heteroskedasticity in the perturbation term, no other assumptions were stated or evaluated for validity. Please do so.

 

There are a few minor issues that can be fixed easily

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

This version may be accepted without any further change as all the points have been thorougly addressed by the authors.

Reviewer 3 Report

Accept in present form.