Neural Networks as an Alternative Tool for Predicting Fossil Fuel Dependency and GHG Production in Transport

Round 1

Reviewer 1 Report

In my opinion, the paper is not novel enough to warrant publication in "Sustainability" in its current form. However, if the author can address my concerns and make major revisions to improve novelty, structure, referencing, and illustrations, the paper may be published in "Sustainability."

Major issues

What is the justification for considering exponential smoothing while there are several linear and non-linear forecasting techniques?

The paper lacks novelty, or I may say the author has not presented the novelty of the paper satisfactorily. For example, in the abstract, the authors have mentioned, "The paper deals with problems of finding new ways for the process of predicting trends in freight transport, especially concerning dependency on fossil fuels and GHG production." But neither exponential smoothing nor ANN are "new ways."

In addition, the author mentions, "Based on publicly available statistical data covering the selected time period, two completely different predicting methods – neural networks and mathematical statistics – are used for forecasting both the above mentioned trends." The same approach of using "neural networks and mathematical statistics" has been used previously in different disciplines. In particular, for carbon emissions the following two articles recently used both "neural networks and mathematical statistics." (1) Machine Learned Artificial Neural Networks Vs Linear Regression : A Case of Chinese Carbon Emissions. doi:10.1088/1755-1315/495/1/012044. And (2) Identifying contributing factors to China's declining share of renewable energy consumption : no silver bullet to decarbonisation. doi:10.1007/s11356-022-20972-x.

It's quite strange for the discussion section to contain so many figures. The conventional discussion section does not contain any figures.

Minor issues

The quality of some figures is inferior and must be improved.

I don't understand the idea behind the restarting for figures numbering from "Figure 1" for each section.

A minimal unit used is "t," which has resulted in an extremely large number of digits in the figures and the tables. I suggest at least using Kiloton [kton] or Megaton [Mton] to improve the readability of the figures and tables.  

Good luck!

Author Response

Uploaded as a PDF file.

Author Response File: Author Response.pdf

Reviewer 2 Report

The subject of the manuscript titled “Neural networks as an alternative tool for predicting fossil 2 fuels dependency and GHG production in transport” is little interesting. It does, however it needs the following revision before moving to the next step.

The abstract can be refined to reflect the key output.

·  State the novelty of the work strongly at end of introduction.

More literature partcan be added in the introduction section. The following articles can be cosidered.

https://doi.org/10.1007/s13399-019-00501-6

https://doi.org/10.1016/j.fuel.2019.02.106

doi.org/10.1080/15435075.2017.1328420

https://doi.org/10.1080/01430750.2016.1181568

https://doi.org/10.1080/01430750.2015.1004106

Figure 5 is not clear. Add clear figure

After figure 5, figure 1 is placed. Proper labelling with correct numbers should be done.

·   All the axis labels are not properly described for the figures.

More discussion needs to be added.

·   The manuscript needs to be proofread by a native speaker thoroughly. A few grammatical mistakes could be found in the manuscript.

Author Response

Uploaded as a PDF file.

Author Response File: Author Response.pdf

Reviewer 3 Report

We appreciate that the paper presents an interesting  analysis;

Some recommendations to authors:

- to explain the acronym GHG

- to make an extended literature review and presentation of the "current requests for sustainable transport and environment both 8 being strongly emphasized in recent EU directions."

- to precise the  research assumptions and the research method,

- to specify the source of the tables, including if it is the author's elaboration

- to highlight the limits of the research

- to specify the usefulness of the research and the theoretical and/or practical contributions

Author Response

Uploaded as a PDF file.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Thanks for the revised article.

Several other software like EViews, MATLAB, R, Phyton, and others develop different types of non-linear regression-based forecasting models. Why do the authors only insist on using the exponential smoothing form the excel still needs to be further justified. 

Secondly, the author should cite some previous articles, especially GHG-related, that also have compared regression and ANN models. 

Author Response

Sent as a PDF file.

Author Response File: Author Response.pdf

Round 3

Reviewer 1 Report

Thanks for the revised version.