1. It is very nice that the processing time is provided to highlight an advantage of the use of a proposed method. It will be better to provide the time for MLE for a comparison to impress readers.
2. It is not clear why PAR dramatically declined using the proposed method compared to the MLE as the number of attributes increases from 3 to 4 and the slipping parameter achieves .5. It seems that there might be something other than the slipping parameter results in this finding.
3. It is not clear to me under what conditions and to what extent the panelized-weightings should be set up.
4. Since this study did not investigate the performance of the proposed method under conditions of small sample sizes, I am curious if it performs as well as the author claims
5. Is it possible to implement discriminant analysis in CDM?