Table 4 Comparisons of primary renal pathological patterns predicted by the three algorithm models on the basis of predicted probability of R2* data
Case Number | Pathological Diagnosis | Predicted by Decision Tree Model (percentage/number, %/n) | Decision Tree Mode Result | Predicted by Line Discriminate Model (percentage/number, %/n) | Line Discriminate Mode Result | Predicted by Logistic Regression Model (percentage/number, %/n) | Logistic Regression Mode Result | |||
|---|---|---|---|---|---|---|---|---|---|---|
III Type | IV Type | III Type | IV Type | III Type | IV Type | |||||
Case 1 | IV | 25% (74) | 75% (226) | IV | 15% (44) | 85% (256) | IV | 3% (8) | 97% (292) | IV |
Case 2 | IV | 42% (125) | 58% (175) | IV | 88% (263) | 12% (37) | III | 56% (176) | 44% (124) | III |
Case 3 | IV | 40% (119) | 60% (181) | IV | 62% (186) | 38% (114) | III | 32% (95) | 78% (205) | IV |
Case 4 | IV | 34% (103) | 66% (197) | IV | 63% (188) | 37% (112) | III | 38% (113) | 62% (187) | IV |
Case 5 | III | 55% (166) | 45% (134) | III | 65% (195) | 35% (105) | III | 37% (112) | 63% (188) | IV |
Case 6 | III | 60% (180) | 40% (120) | III | 81% (243) | 19% (57) | III | 55% (164) | 45% (136) | III |
Case 7 | III | 74% (221) | 26% (79) | III | 88% (263) | 12% (37) | III | 70% (209) | 30% (91) | III |
Case 8 | III | 73% (218) | 27% (82) | III | 65% (194) | 35% (106) | III | 24% (71) | 76% (229) | IV |
Case 9 | III | 58% (173) | 42% (127) | III | 20% (61) | 80% (239) | IV | 5% (14) | 95% (286) | IV |
Case 10 | IV | 23% (70) | 76% (230) | IV | 44% (133) | 56% (167) | IV | 17% (50) | 83% (250) | IV |
Case 11 | IV | 22% (67) | 78% (233) | IV | 1% (4) | 99% (296) | IV | 0% (1) | 100% (299) | IV |
Case 12 | IV | 11% (34) | 89% (266) | IV | 11% (33) | 89% (267) | IV | 1% (4) | 99% (296) | IV |