Evolutionary Multi-Objective Feature Selection Algorithms on Multiple Smart Sustainable Community Indicator Datasets
Round 1
Reviewer 1 Report
Comments and Suggestions for Authors1.English language errors
This paper proposed to perform feature subsets selection and maximizes prediction accuracy on multiple SSC indicators.
NSGA3 outperform Strength...
2. Illegal writing style
subset features -> feaure subsets
Subset feature selection algorithms -> Feature selection algorithms
3. The author proposes Smart Sustainable Community Indicators, and the experiments should use these indicators, not just the number of selected features and classification accuracy.
4. The author needs to use multi-objective evaluation algorithms such as HV, IDG, etc.
5. What are the purposes of ?
Which evolutionary algorithm can select the optimal minimum subset features with high level of accuracy across the multiple smart sustainable city indicators datasets cutting across socio-cultural, economic, environmental and governance? Can smart sustainable city indicators be predicted with minimal subset features?
Comments on the Quality of English LanguageDifficult to understand
Author Response
Please find attached.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThe topic is absolutely interesting and must be studied. I even have a very strong concerns.
1. I'm guessing (because is not clear) that the problem of the work is a problem of selection variables (lines 462 -464) that authors solve by meta-heuristics. The problem is that when meta-heuristics are used, it is because the objective function and constraints are known. In this case, I cannot see any formulation regarding decision variables, constraints or multi-objectives, as stated in line 459 and Fig 1. Thus, I do not know what I am evaluating.
2. According to the above, why use meta-heuristics? Why is your model so complex to solve with exact methods?
3. If the problem is a variable selection one, the author must support your idea of using meta-heuristics to solve this problem. Given that there are other statistical methods to solve it.
4. I have no idea how you get the values in Tables 3 to 8. What exactly is prediction accuracy? The term appears throughout the text, but it is not clear how to compute it.
5. In my opinion, this piece of research lacks reproducibility. I could not reproduce your results to test them.
Comments on the Quality of English LanguageNo comments in this regard.
Author Response
Please find attached
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsNo
Comments on the Quality of English LanguageMinor editing of English language required
Author Response
Thank you very much for your time and recommending my work for publication.
Reviewer 2 Report
Comments and Suggestions for AuthorsDear Authors,
Thank you for your answers. Most of them I agree, wit you response but let me go back to my first comment:
" ... I cannot see any formulation regarding decision variables, constraints or multi-objectives, as stated in line 459 and Fig. 1 .... "
I see through your references, and I notice that in many papers, the formulation is not clearly stated, given that those take for granted that we all know the formulation. It is true, but for the sake of clarity, it would be nice to refer to the mathematical formulation as in [1] and to clarify if your meta-heuristics named in Fig. 2 solve exactly the same problem or if some modifications have been implemented. In the present form, I have no idea of the problem structure to be solved; in my opinion, is like a black box.
[1] Fung, G.M. and Mangasarian, O.L., 2004. A feature selection Newton method for support vector machine classification. Computational optimization and applications, 28, pp.185-202.
Author Response
Please you may wish to find attached.
Author Response File: Author Response.pdf
Round 3
Reviewer 2 Report
Comments and Suggestions for AuthorsA better manuscript.
