Next Article in Journal
Valorizing Organic Waste Through Black Soldier Fly Larvae (Hermetia illucens): A Sustainable Solution for Aquafeeds with Key Nutrients and Natural Bioactive Polyphenols
Previous Article in Journal
Toward Climate-Resilient Freight Systems: Measuring Regional Truck Resilience to Extreme Rainfall via Integrated Flood Demand Modeling
 
 
Article
Peer-Review Record

Machine Learning-Based Carbon Emission Predictions and Customized Reduction Strategies for 30 Chinese Provinces

Sustainability 2025, 17(5), 1786; https://doi.org/10.3390/su17051786
by Siting Hong, Ting Fu and Ming Dai *
Reviewer 1:
Reviewer 2: Anonymous
Reviewer 3:
Sustainability 2025, 17(5), 1786; https://doi.org/10.3390/su17051786
Submission received: 27 January 2025 / Revised: 16 February 2025 / Accepted: 18 February 2025 / Published: 20 February 2025

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript presents a comprehensive approach to predicting carbon emissions and proposing reduction strategies for different provincial categories in China. The authors employ a combination of machine learning algorithms, including an adaptive K-means++ clustering algorithm, GM-SVR model, and SHAP analysis to achieve their objectives. While the study addresses an important and timely topic, several issues need to be addressed to improve the clarity, accuracy, and robustness of the research.

(1) The authors refer to "Low-carbon potential Provinces" (LCPPs) as both "Eco-friendly Provinces" and "Low-carbon potential Provinces" throughout the manuscript. This inconsistency can cause confusion for readers. It is essential to standardize the terminology used for this category to ensure clarity and avoid misunderstandings.

(2) The authors describe the K-means++ algorithm as a classification method, which is incorrect. K-means is a clustering algorithm used for unsupervised learning, not classification. The manuscript should accurately describe the process as clustering the provinces into five distinct clusters rather than classifications.

(3) Table 4 is confusing and difficult to understand the content presented.

(4) While the GM-SVR model demonstrates excellent performance metrics such as low MAE and high R², the authors need to address the potential issue of overfitting. The authors should include additional validation techniques, such as cross-validation or out-of-sample testing, to ensure that the model's performance is robust.

(5) Although the authors mention that the data can be obtained upon request, it is highly recommended to make the dataset publicly available as supplementary material. This practice enhances transparency and allows other researchers to reproduce and validate the results.

Comments on the Quality of English Language

The authors are advised to revise the manuscript for language clarity and readability.

Author Response

非常感谢您审阅我们的稿件。我们已根据您的建议进行了修改,详细信息请参阅附件。

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

The English expression is generally understandable, but there are many formatting errors, and the sentences are verbose and repetitive. Please revise.

The heterogeneity analysis mentioned in the abstract shows the impact of technology and energy consumption on different types of provinces, but the specific mechanisms of these impacts are not detailed. It is recommended to add explanations regarding these impacts in the results section.

The introduction of this article systematically presents the background of global climate change and the importance and challenges of carbon reduction in China. The structure of the introduction is clear, and the content is rich, covering the driving factors of carbon emissions, the limitations of current research, and the proposed new methods. However, there are some writing issues.
In line 4, this paragraph lists various analytical models for carbon reduction, and the author cites a large number of related literature. However, what conclusions can be drawn from these models? What improvements can be made in current research? Has the research question of this paper been distilled from the existing literature review? Please clarify.

In lines 91, 99, 110, 115, and so on, some sentences use (2024) for citations, while others use numerical references at the end of the sentence. Please pay attention to the journal's requirements and check and revise the citation format throughout the text.

In line 143, the name of the figure should be corrected.

In line 120, "co2" should be corrected.

In line 144, this section summarizes the innovations of the paper, which seems to discuss the innovations of the model. However, the introduction should summarize the content of the paper. Please refer to the structure of articles published in other journals.

In the Materials and Methods section, the author only mentions the models used in this paper. What are the data sources? Please supplement this information. Additionally, the research methods lack citations; please add references.

In the results section, you should write the conclusions drawn from your modeling and discuss the result analysis. The data source should be placed in the second part of the Materials section.

Section 3.2 discusses the research methodology and should not be placed in the results section.

In line 29, the author selected four driving factors. Why were only these four chosen? There are many other urban elements that could influence the indicators. What is the basis for this selection?

Figure 2 lacks a scale and a north arrow; please add these. Please pay attention to the norms of figure creation. The figure title is also not standardized; please clarify what (a), (b), (c), and (d) refer to in the title.

Since line 325, the author merely lists which cities belong to which types. Do these types have any distribution characteristics? After completing the data analysis, are there any regular conclusions?

In the analysis of the impact mechanisms, the author chose a series of regression models for modeling, which is a feasible analytical method, but the results of the analysis are not well explained. The author only discusses some parameter indicators of the regression models; please supplement the conclusive analysis of the influencing factors.

Since line 472, many abbreviations appear; please explain what they represent the first time they are mentioned.

In the section on political factors, the specific impacts of these policies are not clearly explained. For example, regarding EGHPS, did the government implement strict energy control strategies that affected carbon emissions? How did the government execute these energy control strategies, and what specific policies were involved? Please provide detailed explanations. Similar issues exist in the discussion of other policies; the author has not explained what policies the government has introduced and how they have brought about impacts.

Comments on the Quality of English Language

The English expression is generally understandable, but there are many formatting errors, and the sentences are verbose and repetitive. Please revise.

Author Response

Thank you very much for reviewing our manuscript. We have made revisions based on your suggestions, and the details can be found in the attached document.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

The review of this paper is mainly based on a methodology of machine learning, using multiple coupling techniques to form a research technical framework, and conducting carbon emission prediction analysis on 30 relevant data points in China. At the same time, PEST's strategy analysis method is used to identify the situation and introduce strategic thinking. The framework design of the methodology in this paper has indicated a novel way of thinking, which is also the significant highlight of this paper. Although the overall architecture is reasonable and the analyzed data is also referenceable. Nevertheless, the writing standards, readability, and presentation of highlights of the article are not ideal, and some errors have appeared in many details, which should be corrected in order to improve the quality and value of the article. The relevant situations and suggestions are as follows:

1. In the keyword section, according to writing standards or principles, the abbreviation in English should be declared throughout the correct full text; otherwise, it may cause misunderstandings with the same abbreviation and reduce readability. This article has experienced multiple instances of this situation; please improve it, such as SHAP (SHapley Additive exPlans), GM-SVR, PEST-SSWOT (such as GM, Lasso, TPE), and EI (expected improvement). In addition, the use of abbreviations in English should be clearly stated at the beginning of the article and should not be abbreviated at the beginning and then suddenly mentioned later in the article (such as GM, Lasso, TPE). Be careful to check and correct it.

2. What is the correct usage of Lasso (least absolute shrinkage and selection operator)? It is used in both uppercase and lowercase in the article, and consistency should be sought.

3. The abbreviation of “Low-carbon Technology-led Provinces (LCTDPs)”. It is an error; please check and correct.

4. In line 95,Sapnken F.E. et al. (2024) proposed an optimized wavelet, The writing is incorrect; it should be rewritten as Sapnken et al. (2024). Please check and correct.

5. In the introduction section of the article, it was found that there was insufficient explanation of the research purpose, and instead, the statement using methodology as the purpose was clearly inappropriate. It is suggested to supplement and rewrite it. In addition, the methodology proposed in paragraphs 131-167 of lines should be described in the Materials and Methods section of Chapter 2. Especially line 141: “In summary, the main contributions of this paper are as follows.” This paragraph should be rewritten as an important highlight of the research methodology rather than a contribution. If the description of the contribution is discussed, it should be appropriately summarized in the conclusion section. In addition, the methodological description mentioned in the conclusion should be presented in the methodology section, which may be more appropriate.

6. Refer to Figure 1 for the description of the sentence 'This is a figure. Scheme follows the same formatting' is clearly incorrect. Please check and correct it. In addition, the role of Projection Pursuit Regression (PPR) should be presented in the technical roadmap of Figure 1. Because this technical method occupies a place in the paper, it should be supplemented to be considered complete and to enhance the readability and understanding of the article.

7. In section “2.1 Self adaptive k-means++algorithm” and “2.2 LASSO Regression,” please supplement mathematical formulas to enhance readability and richness.

8. The labeling of 1. and (1) in the article does not match the general labeling method. In principle, 1 should be used first, followed by the subscript (1). Please check and correct it.

9. The citation errors in the article are as follows:

(1) In lines 98-102, the references [20] should be [01] that annotated by the author, and so on [21] should be [20], [22] should be [21], [24] should be [23], and [25] should be [24]. Additionally, [23] cannot find any references.

(2) 27. 2. Li, X.; Zhang, X. A comparative study of statistical and machine learning models on carbon dioxide emissions prediction of 718 China. Environ. Sci. Pollut. Res 2023, 30, 117485-117502. (Incorrect labeling in the reference 27)

(3) Furthermore, the references have been marked incorrectly since the labeling [20]; thus, the author should carefully proofread the entire document.

10. The Table format in "Table 1. Driving Factor System." is not presented in a standardized manner. Please adjust and strive for aesthetics and neatness.

11. In terms of” Figure 2: the Spatial Distribution Maps of the Four Classification Variables”, the annotations should be made for a, b, c, and d in the map to enhance readability.

12. In the paragraphs 382-383 of the 3.5.1 Driving Factor Prediction section, “the results for the key driving factors of the LCTDPs category in Tianjin from 2022 to 2032 are shown in Table 4.”. The described text seems to be incorrect. I don't know what the key point of expression is. Additionally, what does the mean of TGM in “Table 4 TGM Results for Tianjin”? Meanwhile, what is the significance presented in Table 4? It’s also unable to express clearly in this chapter. Please carefully check and provide additional explanation.

13. In section 3.5.2, Carbon Emission Prediction, there is only a brief description and data chart listed, but the content and key points of the analysis are not stated clearly. Additional explanations should be provided to highlight the value of the analysis. In addition, the resolution of the legend in Figure 8 should be improved.

 

14. In section 4.1 of the paper, the author used the PEST-SWOT analysis method for strategic analysis, but the method seen in the article only uses the PEST analysis method and does not include the content of SWOT. Therefore, it is inappropriate to refer to it as the PEST-SWOT analysis method in the article. Please check and correct it.

Comments on the Quality of English Language

The writing standards, readability, and presentation of highlights of the article are not ideal, and some errors have appeared in many details, which should be corrected in order to improve the quality and value of the article.

Author Response

Thank you very much for reviewing our manuscript. We have made revisions based on your suggestions, and the details can be found in the attached document.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The author has already addressed most of my concerns. However, the 5-Fold Cross Validation should be placed after the introduction of the three evaluation metrics.

Author Response

Thank you very much for reviewing our manuscript. We have made revisions based on your suggestions. Please refer to the attachment for detailed information.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

Thanks for the careful revision, and my concerns have been answered.

Comments on the Quality of English Language

Some minor revision is needed.

Author Response

Thank you very much for reviewing our manuscript. We have made revisions based on your suggestions. Please refer to the attachment for detailed information.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

Upon re-examining the revised content of this article, it was found that the author has made significant and detailed adjustments and revisions to the erroneous content in accordance with previous suggestions, which will help improve the quality and readability of the article. Overall, the article adopts a predictive framework of the GM-SVR model optimized by combining the grey model (GM) and TPE to predict carbon emissions in the next decade, and the research results are validated by the prediction. At the same time, the combination of SHapley Additive exPlanning (SHAP) and PEST-SSWOT strategic analysis methods is used to identify situations and introduce strategic thinking. This article has new ideas in the framework design of methodology, which is also the main highlight of this article. However, there are still two reminders for the author to make slight corrections.

  1. The legend text of the small figures of LCTDPs and LCPPs in Figure 8 is blurry and should be enlarged to improve resolution.
  2. Line 840 LASSO has been changed to Lasso at the beginning of the article; only this part is incorrect. Please correct it.

Comments for author File: Comments.docx

Author Response

Thank you very much for reviewing our manuscript. We have made revisions based on your suggestions. Please refer to the attachment for detailed information.

Author Response File: Author Response.docx

Back to TopTop