Skip to main content

Table 2 Results of leave-one-out cross-validation. Classification accuracy of 169 patients with pleural disease based on LLM and three considered competing methods.

From: Differential diagnosis of pleural mesothelioma using Logic Learning Machine

 

Disease status

 

Classification

MPM

N (%)

MTX

N (%)

BD

N (%)

All

N (%)

Total

Accuracy (%)

LLM

    

77.5

 MPM

41 (78.8)

9 (14.5)

3 (5.5)

53 (31.4)

 

 MTX

6 (11.5)

41 (66.1)

3 (5.5)

50 (29.6)

 

 BD

5 (9.6)

12 (19.4)

49 (89.1)

66 (39.1)

 

DT

    

72.8

 MPM

43 (82.7)

9 (14.5)

5 (9.1)

57 (33.7)

 

 MTX

2 (3.8)

34 (54.8)

4 (7.3)

40 (23.7)

 

 BD

7 (13.5)

19 (30.6)

46 (83.6)

72 (42.6)

 

KNN

    

54.4

 MPM

30 (57.7)

17 (27.4)

7 (12.7)

54 (32.0)

 

 MTX

16 (30.8)

28 (45.2)

14 (25.5)

58 (34.3)

 

 BD

6 (11.5)

17 (27.4)

34 (61.8)

57 (33.7)

 

ANN

    

63.9

 MPM

37 (71.2)

13 (21.0)

12 (21.8)

62 (36.7)

 

 MTX

9 (17.3)

29 (46.8)

1 (1.8)

39 (23.1)

 

 BD

6 (11.5)

20 (32.3)

42 (76.4)

68 (40.2)

 

Total

52

62

55

169

 
  1. MPM = Malignant Pleural Mesothelioma; MTX = Metastasis; BD = Benign Diseases; LLM = Logic Learning Machine; DT = Decision Tree; ANN = Artificial Neural Network; KNN = k-Nearest Neighbour Classifier.