Skip to main content

Table 3 Results for PCA data set. Annotation of the table: Dist: Distribution of senses; S. Size: sample size; Err. Rate: Error Rate; SE: Standard Error of error rates; CV: cross-validation;

From: Machine learning and word sense disambiguation in the biomedical domain: design and evaluation issues

PCA

 

5-fold CV

 

10-fold CV

 

Dist

S. Size

Err. Rate

SE

Err. Rate

SE

(0.5, 0.5)

20

43.00%

12.14%

41.00%

11.25%

 

40

34.58%

8.21%

34.33%

7.68%

 

80

37.17%

5.44%

29.46%

5.14%

 

120

28.53%

4.45%

31.47%

4.13%

(0.6, 0.4)

20

37.83%

11.62%

38.50%

11.04%

 

40

36.42%

8.12%

35.92%

7.37%

 

80

25.54%

5.41%

24.88%

5.06%

 

120

28.22%

4.25%

29.25%

3.96%

(0.7, 0.3)

20

33.67%

11.48%

33.50%

10.90%

 

40

33.08%

8.06%

33.08%

7.62%

 

80

29.67%

5.38%

24.29%

4.98%

 

120

26.83%

4.36%

27.83%

4.11%

(0.8, 0.2)

20

23.67%

10.48%

24.50%

9.99%

 

40

21.83%

7.01%

20.58%

6.61%

 

80

28.00%

5.09%

19.25%

4.61%

 

120

22.92%

3.97%

25.03%

3.65%

(0.9, 0.1)

20

12.33%

8.14%

12.00%

7.59%

 

40

10.92%

5.48%

11.08%

5.20%

 

80

14.04%

4.10%

12.50%

3.79%

 

120

10.42%

3.11%

11.14%

2.98%

(0.67,0.33)

20

38.33%

11.89%

36.33%

10.96%

 

40

30.17%

7.93%

29.50%

7.46%

 

80

28.25%

5.43%

24.83%

5.00%

 

120

29.47%

4.50%

35.33%

4.15%