Using Booleanized Data To Discover Better Relationships Between Metrics.
Susan P. Imberman, Bernard Domanski, Robert A. Orchard · 1999
this paper will be on determining thresholds using the mean (arithmetic average) of data values of each data attribute, the median (middle value) of these data values, the mode (most frequent) of these data values, and by partitioning the values into two groups using clustering techniques. Values above the threshold will take on a Boolean value of 1, and below it a value of 0. The relationships generated by automated methods will be compared to the relationships generated by BA using expert defined thresholds. An analysis of the difference between rule dependency measures will be done using standard statistical analysis