Developing an Optimal Ensemble Model to Estimate Building Demolition Waste Generation Rate
Round 1
Reviewer 1 Report
The submitted paper presents "An optimal ensemble model to estimate building demolition waste generation rate". Although the paper might have some novelties, some points need clarification:
1) If the main purpose bias reduction, LOOCV should be preffered to k-fold CV since it tends to has less bias. But in GBDT each estimator(base learner) should have high bais so that it can increase bais by boosting. Justify
2) What about other hyperparameters like row sampling rate , column sampling rate, depth of tree, shrikage(learning rate)
3) Why you have not oob points for CV ?
4) You have used RSME, R as performance parameters. Why not other parameters are employed. Justify
5) Why only DT why not SVM or any other machine learning algorithim ?
Author Response
Please see the attachment
Author Response File: 
Author Response.docx
Reviewer 2 Report
Overall paper is good and interesting work. Some of the minor observations are as follow:-
1. Abstract is vague
2. What are the various limitations in your study ?
3. Give citation wherever required
4. Why author have selected only arid areas?
5. How many governing equations are involved?
6. Any assumptions considered in your study?
7. How many algorithms you have used?
8. Have you validate your model ?
9. Which machine is more effective ?
Author Response
Please see the attachment
Author Response File: Author Response.docx
Reviewer 3 Report
The manuscript presented offer a novel study on the theme of demolition waste and the recent application of machine learning. However, some suggestions and grammar checks need to be fix before publication.
The abstract may be redesigned in the light of its connection with methodology and limitations of the study. Abstract should cover the critical findings of the study at all.
In the introduction section, the long part of the description of R2 with several numbers can be summirazed in a table of these contents.
Citation references can be written in another way. For example, Noori et al. (2009): the year of publication is not required in the text.
Lines 112-124 provide valuable insights into the general concept of machine learning and the necessary and useful data for the present study. I recommend moving this paragraph before the lengthy portion in the introduction, which begins at line 52.
Lines 133 authors can add the number of introduction section (section 1).
Lines 136- 137 “and, Conclusions, together with the application plan and 136 limitations of the study are presented in Section 5”. This sentence can be written in a better way.
Line 141 “redevelopment target” need a English check.
Line 158 “2.2. Data preprocessing and dataset size” some information more can be added regarding the dataset choice and accurate numerical process done.
Line 239 “2.5.1. Model validation” Are you sure it is the correct name for the sub paragraph? In the 2.5 it is written evaluation. maybe authors can improve this section in a more accurate way.
Line 261 Example of citation that has to be improved: “…defined in Eq. (8; [60])”. Please check mistakes all over the text.
Figure 1 and Figure 2 need to be arranged in four different figures (Figure 1a, Figure 1b., etc)
Line 357 which are Figure A1 and A4? Please check all the figure and table numbers.
Generally discussion and conclusion sections need connection points.
Line 378 Figure 5 need a layout improvement.
Lines 394-397 Authors maybe can show the percentage rate in a table or in a graph display.
Lines 451 -455 can we call this future development?
Author Response
Please see the attachment
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
My comments have been addressed. The manuscript can now be accepted.
Reviewer 2 Report
The revised manuscript quite acceptable
Reviewer 3 Report
The changes made meet the revision requests. In this form, the article can be accepted.
