Solid Waste Analysis Using Open-Access Socio-Economic Data

Round 1

Reviewer 1 Report

Figure 1 & 2 are Tables.

Table description should be given above the table.

Related work needs improvement.

Conclusion should be there.

Some more Latest references required.

 

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

This paper presents a novel method with an attractive application, which probably can be published and gain future research attention. Therefore, I recommend the authors some advice to improve the quality of this paper, which is as follows.
After reading the paper and the abstract can be improved to deliver the main idea of the research for those unfamiliar with this domain easily. Also, the details of the conducted contributions, experiments, comparisons, and obtained results should be added into the abstract section properly.
The introduction section can be presented in another way. The authors can give the introduction section in some terms like the general idea that already demonstrates an interest these days, the particular domain that uses the current application, the main problem in this research, and the central gap founded by the authors in this paper. Some related works support the main claim in this paper and support this work by focusing on some issues. At the end of the introduction section, the authors should be given a clear and comprehensive paragraph to show the readers how this research has been done (the main problem, contributions, experiments, comparisons, results, and so on).
I feel that the flow of the proposed method can be improved as well the mathematical notations also need to be checked and revised.
The authors should give the readers a straightforward method of choosing the experiments and their design. This will help the future researcher in this domain conduct new research and start from the current paper.
For example, the following papers might be cited in your work. Reptile Search Algorithm (RSA): A nature-inspired meta-heuristic optimizer
Feature selection and enhanced krill herd algorithm for text document clustering
The discussion of the results needs to include the strengths and weaknesses of the proposed algorithm.
Furthermore, where are the limitations of your study? Clarifying the study's limitations allows the readers to better understand under which conditions the results should be interpreted. A clear description of the limitations of a study also shows that the researcher has a holistic understanding of his/her study. However, the authors fail to demonstrate this in their paper.

%%%%%
The organization of this paper should be enhanced.
Some recent works from high-quality journals should be cited in this research.

summarize the related works to be more beneficial for the readers

I suggest removing the abbreviation from the title

 

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 3 Report

Solid Waste Analysis using Open Access Socio-Economic Data from the OECD.

Review reports

Summary:

The paper looks into the OECD data on waste and uses machine learning methods to analyse waste production of OECD countries. Particularly, the work clusters countries according to their waste and socio-economic features, identify key features that can be used for prediction of waste production, and comparison of several ML methods for the tasks. The authors conclude that the Random Forest Regressor produces the best prediction results over the dataset.

General Comments:

There are a few grammatical and structural errors in the manuscript, e.g 193-194, line 803. Also, instead of referring to the problem as waste consumption in the manuscript, e.g. line 182 and others, it should be waste production.

Specific comments

 

1. Abstract

• The paper has highlighted 3 contributions from the work in the abstract, however, have not mentioned specific results in the abstract, e.g. what are the most relevant features, results of their 1^st contribution, etc.
• Please do not abbreviate OECD on first use, in the abstract
• I would like to see in the abstract the significance of the work, and how it can be useful?

 

2. Introduction

• Line 32: “In consequence, ..” should be consequently.

3. Related Work

• Other papers in the literature have also made similar attempts to predict waste production, e.g line 78. I would like to see how successful and how accurate they are in their prediction, as well as the data that they used for their prediction works. The results of others should form the basis for comparison for the success of the prediction method proposed in this paper

4. OECD dataset

• Line 182: “..that waste consumption..” should be waste production
• Line 192-193: English error
• Figure 3.: Can you comment on the correlations between built area/population vs age structure as there seem to be some strong correlations
• Figure 3: I would also say that Income and Secondary education have a moderate negative correlation

5. Data Analytic methods

• I would suggest that you sub-section Section 4, with Section 4.1: Selected ML algorithms (consisting of sub-sub sections of the different algorithms considered), and Section 4.2: Evaluation metrics.
• For the different ML algorithms, can you please provide some references for these methods
• For The evaluation metrics, can you please provide interpretation of the measures, e.g. whether high or low MAPE is better, the difference between the measures and which ones is the most siginificant and would be used as an important indicator. From your result, you mostly used MAPE only.
• Nowadays, it is common to provide source code for your algorithms as well. This can be put in a github repository

6. Analysis of the OECD data

• Table 4: there are 4 measures but in lines 488-472, you only concentrated on MAPE. This ties back to my previous comment. Why is MAPE more important, and if the rest of the measures are not even worth commenting, you may want to remove!
• Figure 6: Please label the x and y axis properly.
• Line 495: I would like to know how the feature importance in Figure 7 are calculated, and its interpretation. Maybe this can be included in the section on Evaluation Metrics.
• Figure 7: please label the axis properly
• Line819-827: I would like to see this comparison of different Machine learning methods appear at the beginning of the result on this section, to form a basis of why random forest is used for the rest of the section. Also, why only MAPE is used as a measure, and not other metrics?
• In Section 5.2.2, there are no comparison of different ML algorithms. Although Random forest is used optimum for section 5.2.1, it may not be the cased for the prediction of waste from other waste types. I would like to see this comparison here.
• Figure 8: please put (a) and (b) as well as label the x and y axis properly.
• Line 863: I would not say the feature importance corresponds to the results of your co-variance analysis in Section 3.3 in its entirety; e.g. the co-variance between municipal and recover is 0.76 in figure 4 (in fact the lowest), but it is one of the highest in Figure 9 (a), and for household waste, its co-variance with compost I quite high in Figure 4 (0.88), but almost 0 in Figure 9 (b). You need to provide the explanations for these discrpenacies.
• Figure 9: please label and put Figure (a), (b), etc.
• Line 862: similar to previous section, I would like to know how the feature importance in Figure 9 are calculated, and its interpretation. Maybe this can be included in the section on Evaluation Metrics.
• Finally, I would like comparison of your method with other methods available in the literature.

7. Discussion

• Maybe can be changed to Conclusion
• Highlight the main contributions, and main results in this section please. This section is quite weak
• I also like to see more references. 25% of the references are more than 10 years old; the paper needs to include more recent references.

Comments for author File: Comments.pdf

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Accept in present form

Author Response

Thank you again very much for your constructive comments, which have helped us to improve the new version of the manuscript. 

Reviewer 3 Report

New Comment 1:

In the Abstract: “Effective waste management systems at different geographic levels need from accurate forecasting…” can change to “Effective waste management systems at different geographic levels requires accurate forecasting…”

 

From previous comment:

- Figure 3 : Can you comment on the correlations between built area/population vs age structure as there seem to be some strong correlations

We have added some more explanations in section 3: It can also be observed that the POPULATION and BUILTAREA correlate with the age structure. This holds, because all these features correspond to the size of a country. If the population or the built-up area of a country is large, then the absolute number of people of a certain age is also large.

New comment:-

I would say population is strongly correlated with Under 20, Over 50 and Over 65; but for built-area, it is correlated strongly with Over 85.

 

Author Response

Thank you very much for your  comments! 

We have changed the sentance in the abstract and added your comment to section 3.