Nonparametric method for measuring clustered level of time rank in binary data

1. A nonparametric method for measuring a clustered level of time rank in binary data, the nonparametric method comprising:

classifying a sample set of engineering data into a target group and a reference group;

setting a rank to each sample of the sample set in a chronological order;

obtaining a minimum rank and a maximum rank from the target group and setting a characteristic period based on the minimum rank and the maximum rank, and incorporating the sample set included in the characteristic period into an analysis data set;

calculating an average rank value of the target group and an average rank value of the reference group in the characteristic period;

creating a dummy sample set, wherein a total of the number of samples of the dummy sample set and the number of the samples classified into the reference group in the characteristic period is equal to the number of the samples classified into the reference group in the engineering data;

incorporating the dummy sample set into the analysis data set and setting a new rank to each sample of the analysis data set, based on a comparison result of the average rank value of the target group and the average rank value of the reference group and the minimum rank and the maximum rank of the characteristic period, to obtain adjusted data; and

executing a Mann-Whitney U test on the adjusted data to obtain a clustered level index.

2. The nonparametric method according to claim 1 , wherein after the step of calculating the average rank value of the target group and the average rank value of the reference group in the characteristic period, the nonparametric method further comprises:

comparing the average rank value of the target group with the average rank value of the reference group to obtain the comparison result, so as to decide an order of the dummy sample set and the sample set included in the characteristic period based on the comparison result when incorporating the dummy sample set into the analysis data set.

3. The nonparametric method according to claim 2 , wherein the step of incorporating the dummy sample set into the analysis data set and setting the new rank to each sample of the analysis data set comprises:

reversing the ranks of the sample set included in the characteristic period if the average rank value of the target group is greater than the average rank value of the reference group; and

setting the dummy sample set to succeed the maximum rank of the reversed characteristic period and setting new ranks to obtain the adjusted data, wherein each of new ranks of the dummy sample set in the adjusted data is obtained by adding the maximum rank of the reversed characteristic period and each of initial rank of the dummy sample.

4. The nonparametric method according to claim 2 , wherein the step of incorporating the dummy sample set into the analysis data set and setting the new rank to each sample of the analysis data set comprises:

if the average rank value of the target group is greater than the average rank value of the reference group, setting the dummy sample set to be prior to the minimum rank of the characteristic period and setting new ranks to obtain the adjusted data, wherein new ranks of the dummy sample set in the adjusted data are set from 1, and new ranks of the characteristic period are set from a value obtained by adding 1 and maximum rank of dummy sample set in the adjusted data.

5. The nonparametric method according to claim 2 , wherein the step of incorporating the dummy sample set into the analysis data set and setting the new rank to each sample of the analysis data set comprises:

if the average rank value of the target group is smaller than the average rank value of the reference group, setting the dummy sample set to succeed the maximum rank of the characteristic period and setting new ranks to obtain the adjusted data, wherein each of new ranks of the dummy sample set in the adjusted data is obtained by adding the maximum rank of the characteristic period and each of initial rank of the dummy sample.

6. The nonparametric method according to claim 2 , wherein the step of incorporating the dummy sample set into the analysis data set and setting the new rank to each sample of the analysis data set comprises:

if the average rank value of the target group is equal to the average rank value of the reference group, setting the dummy sample set to succeed the maximum rank of the characteristic period and setting new ranks to obtain the adjusted data, wherein each of new ranks of the dummy sample set in the adjusted data is obtained by adding the maximum rank of the characteristic period and each of initial rank of the dummy sample, or setting the dummy sample set to be prior to the minimum rank of the characteristic period and setting new ranks to obtain the adjusted data, wherein new ranks of the dummy sample set in the adjusted data are set from 1, and new ranks of the reversed characteristic period are set from maximum rank of dummy sample set in the adjusted data plus 1.

7. The nonparametric method according to claim 1 , wherein after the step of setting the characteristic period based on the minimum rank and the maximum rank, the nonparametric method further comprises:

if the characteristic period does not include the sample that belongs to the reference group, setting the dummy sample set to succeed the maximum rank of the characteristic period and setting new ranks to obtain the adjusted data, wherein each of new ranks of the dummy sample set in the adjusted data is obtained by adding the maximum rank of the characteristic period and each of initial rank of the dummy sample, or setting the dummy sample set to be prior to the minimum rank of the characteristic period and setting new ranks to obtain the adjusted data, wherein new ranks of the dummy sample set in the adjusted data are set from 1, and new ranks of the characteristic period are set from a value obtained by adding 1 and maximum rank of dummy sample set in the adjusted data.