An FDA-Based Approach for Clustering Elicited Expert Knowledge
, Juan Carlos Correa 2, Andrew Zamecnik 3, Francisco Torres-Avilés 4,*,† and Fernando Marmolejo-Ramos 3,*Round 1
Reviewer 1 Report
The article presents an interesting approach to solving the segmentation of expert knowledge. However, the text needs to be improved.
It is necessary to highlight the motivation and reason for using these specific approaches. It especially compared with other methods as multicriteria decision-making of fuzzy.
The Conclusion section is missing. The Discussion section contains the Conclusion content instead. In the Discussion, include research gaps and Future development.
Correct the minor formal and stylistic errors.
Figure 1 contains a border on the right. For this type of pictures needs to note the authorship.
Author Response
see attachment
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Please see attached comments.
Comments for author File: Comments.pdf
Author Response
see attachment
Author Response File: Author Response.pdf
Reviewer 3 Report
The review is attached.
Comments for author File: Comments.pdf
Author Response
see attachment
Author Response File: Author Response.pdf
Reviewer 4 Report
In this manuscript, the authors propose a method for clustering elicited distributions. The performance of the novel method is demonstrated via an extensive empirical study. The authors do not provide theoretical results on the properties of the newly proposed method.
This reviewer considers that the following points should be addressed in the improved version of the manuscript:
- Page 5: The authors write: “Discretize the curves using a grid of m points (y1,…,ym) These points correspond to heights of the curve in m different points of the curve space (we used 100 points).” It is not clear how the number 100 was selected and it is not clear how the 100 points are selected for each curve. The authors should provide complete explanations as these selections have an important effect on the outcome of the algorithm.
- Page 7: The authors calculate “the Rand index (R), the Jaccard coefficient (J), the Fowlkes-Mallows index (FM)”. However, it is not clear why the Adjusted Rand index (when the expectation of the index takes some constant value under an appropriate null model for the contingency table) is not used. The authors should clarify this aspect.
See the following references:
Hubert and P. Arabie, “Comparing partitions, ”J. Classification, vol. 2, pp. 193–218, 1985.
Morey and A. Agresti, “The measurement of classification agreement: an adjustment to the Rand statistic for the chance agreement,”Educ. Psychol. Meas., vol. 44, pp. 33–37, 1984
3. Figure 6: The significance of the colors is unclear.
Author Response
see document
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Thank you for correcting the text in the sense of the sent recommendations. The text is now of a higher quality and can therefore be recommended for acceptance.
Author Response
Thank you very much for your comments.
Reviewer 2 Report
I am satisfied with the edits of the paper
Author Response
Thank you very much for your comments.
