Table 1 Added data counts and accuracies after each iteration of semi-supervised learning.

From: Evaluation of two semi-supervised learning methods and their combination for automatic classification of bone marrow cells

Times of semi-supervised learning

CST + AL

CST

AL

Accuracy_CST + AL

DATA_Count_CST + AL

Accuracy_CST

DATA_Count_CST

Accuracy_AL

DATA_Count_AL

1

0.83500

425

0.82250

425

0.80875

425

2

0.85000

1694

0.86000

1603

0.81375

828

3

0.86850

3298

0.85500

2719

0.83125

876

4

0.90125

4881

0.87250

3985

0.87375

1150

5

0.91000

6236

0.88625

5487

0.87750

1227

6

0.93125

7901

0.89125

7187

0.90000

1355

7

0.92875

8783

0.86875

8852

0.88000

1597

8

0.93750

10,804

0.90125

10,082

0.90750

1701

9

0.93625

11,981

0.89375

11,572

0.91875

1814

10

0.94000

13,624

0.89750

13,332

0.91375

1909

11

0.92875

14,907

0.88750

15,079

0.91750

1963

12

0.93250

16,966

0.92250

16,619

0.92125

2172

13

0.94125

19,558

0.89500

18,227

0.91125

2250

14

0.94750

21,903

0.90875

19,889

0.92000

2420

15

0.95375

23,774

0.89375

21,470

0.93000

2593

16

0.96000

26,231

0.91375

23,284

0.92250

2711

17

0.95875

31,603

0.91750

25,163

0.92000

2791

18

0.95750

33,787

0.92000

26,997

0.90875

2843

19

0.95625

36,585

0.91500

28,886

0.91875

2974

20

0.97500

38,465

0.92625

30,930

0.94000

3073

21

0.96625

40,778

0.91750

32,493

0.94125

3178

22

0.96875

43,232

0.93750

34,766

0.94125

3304

23

0.96125

45,112

0.93000

36,761

0.93250

3395

24

0.97250

46,006

0.93750

38,051

0.93000

3574

25

0.97625

47,843

0.94375

40,518

0.93500

3682