Skip to main content

Table 3 Classification error rates (%) at different concentrations using Linear (L), Mahalanobis (M), and Quadratic (Q) type DA methods.

From: Classification of root canal microorganisms using electronic-nose and discriminant analysis

 

Concentration

Dataset (pre-processing & dimension reduction)

12 × 10 8 cfu/ml

12 × 10 5 cfu/ml

12 × 10 3 cfu/ml

12 × 10 1 cfu/ml

 

L

M

Q

L

M

Q

L

M

Q

L

M

Q

d1 & std

14.29

2.86

0.00

25.71

5.71

0.00

31.43

5.71

2.86

20.00

8.57

2.86

d1 & pca

5.71

0.00

0.00

25.71

5.71

2.86

22.86

5.71

0.00

17.14

11.43

5.71

fd1 & std

17.14

8.57

2.86

25.71

2.86

5.71

25.71

5.71

2.86

5.71

2.86

0.00

fd1 & pca

5.71

2.86

0.00

17.14

0.00

0.00

34.29

8.57

8.57

20.00

5.71

0.00

nfd1 & std

11.43

17.14

0.00

0.00

0.00

0.00

22.86

8.57

5.71

8.57

2.86

5.71

nfd1 & pca

11.43

17.14

0.00

0.00

0.00

0.00

22.86

2.86

2.86

11.43

5.71

2.86

d2 & std

34.29

8.57

2.86

17.14

8.57

11.43

22.86

5.71

8.57

25.71

17.14

11.43

d2 & pca

31.43

0.00

0.00

17.14

8.57

11.43

31.43

14.29

8.57

34.29

28.57

22.86

fd2 & std

17.14

2.86

0.00

17.14

0.00

0.00

37.14

14.29

17.14

20.00

14.29

14.29

fd2 & pca

20.00

2.86

2.86

14.29

5.71

2.86

34.29

2.86

2.86

25.71

8.57

2.86

nfd2 & std

14.29

5.71

5.71

8.57

2.86

2.86

22.86

2.86

2.86

5.71

2.86

2.86

nfd2 & pca

14.29

8.57

0.00

8.57

2.86

0.00

14.29

8.57

5.71

14.29

8.57

2.86

d3 & std

20.00

5.71

0.00

28.57

0.00

2.86

28.57

8.57

2.86

25.71

14.29

5.71

d3 & pca

8.57

0.00

0.00

31.43

5.71

0.00

25.71

8.57

2.86

17.14

11.43

11.43

fd3& std

14.29

2.86

0.00

20.00

0.00

2.86

28.57

11.43

5.71

14.29

5.71

5.71

fd3 & pca

5.71

5.71

0.00

17.14

2.86

0.00

22.86

0.00

0.00

11.43

5.71

8.57

nfd3 & std

11.43

8.57

2.86

2.86

2.86

0.00

20.00

8.57

8.57

5.71

2.86

2.86

nfd3 & pca

11.43

8.57

0.00

5.71

5.71

5.71

14.29

8.57

5.71

8.57

2.86

0.00

  1. These datasets were obtained using different pre-processing and dimension reduction methods.
  2. (For the abbreviations used in the table, see Table 2.)