Fascinating how you formulated the problem. I can see lot of use if this in actuarial analyses of large sets of data, where possibly being right about every individual outcome is not as important as highlighting interesting subgrouos within the data set with high risk of, in this case, diabetes.
Fascinating how you formulated the problem. I can see lot of use if this in actuarial analyses of large sets of data, where possibly being right about every individual outcome is not as important as highlighting interesting subgrouos within the data set with high risk of, in this case, diabetes.