Skip to main content

Advertisement

Table 8 AUC Results for Train_CV

From: The effects of class rarity on the evaluation of supervised healthcare fraud detection models

Learner Ratio 200 400 1000 All
(a) Part B
GBT [Full] 0.75982 0.78328 0.79120 0.79569
[1:99] 0.76740 0.79520 0.80378 0.80373
[10:90] 0.76032 0.79377 0.81847 0.82064
[25:75] 0.74964 0.78624 0.81464 0.81948
[35:65] 0.73271 0.77326 0.80600 0.81434
[50:50] 0.71530 0.75244 0.79563 0.80499
LR [Full] 0.77162 0.78921 0.80019 0.80516
[1:99] 0.78282 0.79295 0.81119 0.81238
[10:90] 0.77752 0.79680 0.81465 0.81881
[25:75] 0.76797 0.79746 0.81507 0.81686
[35:65] 0.75771 0.79061 0.81336 0.81806
[50:50] 0.73414 0.78012 0.80964 0.81415
RF [Full] 0.71510 0.73806 0.78110 0.79604
[1:99] 0.74846 0.77197 0.80579 0.81586
[10:90] 0.76661 0.79117 0.81933 0.83012
[25:75] 0.76031 0.79187 0.81641 0.82703
[35:65] 0.75699 0.78061 0.81299 0.82156
[50:50] 0.74994 0.77298 0.80448 0.81496
Learner Ratio 100 200 400 All
(b) Part D
GBT [Full] 0.68101 0.71044 0.73932 0.74851
[1:99] 0.68871 0.70412 0.74731 0.75727
[10:90] 0.66033 0.69299 0.74381 0.76756
[25:75] 0.65692 0.67700 0.73008 0.76538
[35:65] 0.63219 0.66694 0.71228 0.75996
[50:50] 0.62040 0.65461 0.70773 0.74506
LR [Full] 0.72516 0.75436 0.77369 0.78164
[1:99] 0.71200 0.75396 0.77575 0.78486
[10:90] 0.71031 0.75129 0.77481 0.78657
[25:75] 0.70331 0.73115 0.77009 0.78540
[35:65] 0.67880 0.72835 0.76340 0.78216
[50:50] 0.67158 0.70696 0.74834 0.77557
RF [Full] 0.62721 0.63364 0.66818 0.70888
[1:99] 0.67627 0.68215 0.70816 0.73706
[10:90] 0.67777 0.69735 0.73538 0.75857
[25:75] 0.67634 0.69916 0.72832 0.75838
[35:65] 0.65126 0.68992 0.72510 0.74904
[50:50] 0.64951 0.68343 0.70771 0.74088
Learner Ratio 100 200 400 All
(c) DMEPOS
GBT [Full] 0.67203 0.68827 0.72125 0.73129
[1:99] 0.66654 0.68611 0.72516 0.73591
[10:90] 0.65411 0.68073 0.72241 0.73777
[25:75] 0.64571 0.66342 0.71327 0.73389
[35:65] 0.61468 0.64740 0.70118 0.72090
[50:50] 0.60699 0.63259 0.68728 0.70598
LR [Full] 0.68783 0.70311 0.73615 0.74063
[1:99] 0.68960 0.69565 0.73853 0.74085
[10:90] 0.67423 0.69604 0.73498 0.74421
[25:75] 0.66667 0.68769 0.72912 0.73715
[35:65] 0.65088 0.68259 0.72463 0.73488
[50:50] 0.64590 0.66432 0.71445 0.72225
RF [Full] 0.61229 0.64745 0.69381 0.70756
[1:99] 0.64896 0.66998 0.70598 0.72245
[10:90] 0.65829 0.67671 0.72066 0.73767
[25:75] 0.65636 0.67337 0.71790 0.72889
[35:65] 0.64239 0.67054 0.71756 0.72390
[50:50] 0.63938 0.66152 0.70306 0.72379
Learner Ratio 100 200 All  
(d) Combined
GBT [Full] 0.73906 0.76623 0.79047  
[1:99] 0.73626 0.78562 0.80373  
[10:90] 0.72482 0.76730 0.81675  
[25:75] 0.68806 0.75833 0.80405  
[35:65] 0.68275 0.74855 0.79127  
[50:50] 0.65960 0.72675 0.77587  
LR [Full] 0.74260 0.80043 0.81554  
[1:99] 0.73814 0.80060 0.82011  
[10:90] 0.72508 0.78653 0.81868  
[25:75] 0.69117 0.77479 0.81553  
[35:65] 0.67940 0.76854 0.80998  
[50:50] 0.67567 0.74588 0.79415  
RF [Full] 0.64769 0.71098 0.79383  
[1:99] 0.71813 0.76663 0.81515  
[10:90] 0.73110 0.79011 0.82793  
[25:75] 0.74162 0.77822 0.81503  
[35:65] 0.72834 0.76699 0.80619  
[50:50] 0.71446 0.76228 0.79546