A Review of Macroscopic Carbon Emission Prediction Model Based on Machine Learning

Round 1

Reviewer 1 Report

This paper extensively reviews the application of various machine learning algorithms in predicting future carbon emissions, comparing the characteristics of popular algorithms such as the BP neural network, support vector machine, long short-term memory neural network, random forest, and extreme learning machine. The paper contributes to understanding how machine learning algorithms are applied in carbon emission predictions and suggests potential research directions for their application.

 

This paper offers a comprehensive review of machine learning algorithms and their use in research. To improve, it could provide trends in the application of machine learning algorithms for carbon emission predictions. Some algorithms have been used for many years, while others may have been recently developed and focused on in recent years. Although some information is provided in the manuscript, a clear presentation of trends could help readers discern which algorithms might become more popular in the future.

 

Additionally, the improvement in accuracy and efficiency, as well as the relaxation of strict assumptions, are strengths of carbon emission predictions using machine learning algorithms compared to traditional econometric methods. However, the traditional methods explain more of the key factors determining carbon emissions. The study's contribution could be enhanced by emphasizing the strengths of machine learning algorithms relative to econometric methods and discussing how each algorithm can overcome the challenges of weak explanatory power for each factor or variable.

 

Furthermore, the type and frequency of the dataset could be important factors in selecting machine learning algorithms, given their distinct characteristics. In recent years, as more data and information become available, their use in carbon emission predictions will become increasingly important. For example, in addition to national-level information such as GDP, data on firm-level or household-level characteristics could be considered when predicting carbon emissions and incorporated into machine learning algorithms.

 

Lastly, while this paper provides a solid review and comparison of algorithms, its discussion of application prospects is relatively weak.

Author Response

Please see the attachment.（Due to fine-tuning of the manuscript format, the number of lines has changed）

Author Response File: Author Response.docx

Reviewer 2 Report

This study presents a literatüre survey of carbon emission prediction models based on machine learning. My suggestion for the paper is given below;

 

- The abstract must not include any abbreviation.

- All figures must be given before their description in the text (example of figure 1).

--The manuscript was written in pure language but it needs still syllable and grammar checks.

- it will be useful for the reader if traditional and machine learning-based carbon emission prediction methods are compared by giving examples of studies in the literature in a table.

-The meanings of the symbols in all equations should be given either in the text or in a nomenclature.

- Is the location of the abbreviations table correct?

- All table, figure, and reference styles must be suitable to journal format.

- Although it was mentioned in the abstract section that future work suggestions were given, I can not find this in the manuscript. Future works recommendations must discuss in a separate title and future works recommendations for researchers must be listed in detail.

- Authors should examine the literature in more detail.

- The study titles and the structure of the study can be revised. For example, future works may be reviewed under a separate heading and a summary of study findings may be presented in the conclusion.

- In the reference table, there is some typos, these must correct.

- it must detaily highland the what are contributions of this study in the introduction section.

Author Response

Please see the attachment.（Due to fine-tuning of the manuscript format, the number of lines has changed）

Author Response File: Author Response.docx

Reviewer 3 Report

In this study, the current situation of machine learning applied to carbon emission prediction is studied in detail by means of paper retrieval. It is found that machine learning has become a hot topic in the field of carbon emission prediction model and the main carbon emission prediction models are mainly based on BP neural network, support vector machine, long short-term memory neural network, random forest and extreme learning machine. In general, this study has important implications for the study of carbon emission, The authors have made a comprehensive summary of the literature, and I think it can be published after undergoing minor revision.

1. In the introduction section, it is suggested that the authors appropriately add a background statement on the severity of carbon emissions to further highlight the significance of this study. Meanwhile, it is suggested to appropriately increase the number of references to the seriousness of the carbon emission problem. The following literature is for reference (Seneviratne et al., 2016; Zhang et al., 2019; Song et al., 2022).

2. It is suggested that the authors make clear the contribution of this study to the field in the introduction.

3. 5. Summary and application prospect, it is suggested that the authors further point out which problems of carbon emission can be studied by the machine learning prediction model in the future, so as to point out the direction for future research.

 

Seneviratne, S.I., Donat, M.G., Pitman, A.J., Knutti, R., Wilby, R.L., 2016. Allowable CO2 emissions based on regional and impact-related climate targets. Nature 529 (7587), 477–483.

Zhang, C., Su, B., Zhou, K., Yang, S., 2019. Decomposition analysis of China’s CO2 emissions (2000–2016) and scenario analysis of its carbon intensity targets in 2020 and 2030. Sci. Total Environ. 668, 432–442.

Song, Y., Zhang, Y., Zhang, Y., 2022. Economic and environmental influences of resource tax: Firm-level evidence from China. Resources Policy, 77, 102751.

Author Response

Please see the attachment.（Due to fine-tuning of the manuscript format, the number of lines has changed）

Author Response File: Author Response.docx

Reviewer 4 Report

The presented research entitled A Review of Macroscopic Carbon Emission Prediction Model Based on Machine Learning is a review of professional scientific literature, published on the topic in impact journals and indexed on WOS.

Quite logically, the main scientific method here is the search of professional literature and the subsequent statistics and presentation of the obtained results.

Authors with a sufficient amount of literary sources worked for this purpose.

It can be stated that the obtained results are rather interesting.

However, as a weaker part, I have to mention some errors in the presentation and construction of the article.

- Specifically, all tables, figures and graphs lack citations. It should be remembered that even if it is probably an author's invention, it is necessary to indicate as a source, for example, one's own creation.

- Furthermore, Figures 1 lacks axis markings in the graph

- Figures 2 is illegible and needs to be enlarged.

- The formula on line 101 does not have variables described in the formula after the where interjection:

- Table 1 is inappropriately divided, when there is only the title and one row on the 1st page

- Figure 3 is also completely illegible and needs to be enlarged

- Also, the formula on line 207 does not have the subsequent variables described

- Table 3 is also inappropriately divided into 3 pages, where on the 1st page there is again only an inscription and one line

- The title of subsections 3.3 and 3.3.1 is completely blank at the end of the page, which is unacceptable for publication

- Figures 7 and 10 are also completely illegible and need to be enlarged

- Furthermore, I must mention that the final chapter of the Summary is relatively well summarized in 5 basic points of findings, however, the introduction chapter does not correspond with this.

Here I would suggest that the authors insert a section in the Introduction section where the research objectives and questions will be clearly stated, which would correspond to the 5 points in the Summary section.

Author Response

Please see the attachment.（Due to fine-tuning of the manuscript format, the number of lines has changed）

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Dear Authors, 

Thank you for your efforts in responding to my review comments on your paper, and particularly, the 'prospect' section has been strengthened, and the contribution of the paper seems more clear and more visible.

Thank you

Reviewer 2 Report

As future works suggestions and study results are evaluated in Chapter 6, the title of this chapter can be revised appropriately.

After be controlled in terms of english grammar and syllable, the manuscript can be accept.

Reviewer 4 Report

The article is improved by revision on the level needed for the publication of the journal.