Skip to main content

Table 4 Confusion matrix of the best system (feature set ℬ MathType@MTEF@5@5@+=feaagaart1ev2aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacPC6xNi=xH8viVGI8Gi=hEeeu0xXdbba9frFj0xb9qqpG0dXdb9aspeI8k8fiI+fsY=rqGqVepae9pg0db9vqaiVgFr0xfr=xfr=xc9adbaqaaeGaciGaaiaabeqaaeqabiWaaaGcbaWenfgDOvwBHrxAJfwnHbqeg0uy0HwzTfgDPnwy1aaceaGae8hlHieaaa@36A6@ , SDA).

From: An incremental approach to automated protein localisation

  

classification results (in percent)

cell compartment

 

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

(j)

(*)

cytoplasm + nucl.

(a)

97.2

2.1

0.0

0.7

0.0

0.0

0.0

0.0

0.0

0.0

0.0

cytoplasm – nucl.

(b)

17.9

75.0

1.8

0.0

3.6

0.0

0.0

0.0

0.0

1.8

0.0

ER

(c)

0.0

0.0

81.0

2.8

7.0

7.0

0.7

1.4

0.0

0.0

0.0

lysosomes

(d)

0.9

0.5

3.6

85.6

0.0

7.7

0.9

0.5

0.0

0.5

0.0

microtubules

(e)

0.0

0.0

9.8

1.0

87.3

2.0

0.0

0.0

0.0

0.0

0.0

mitochondria

(f)

0.0

0.0

4.1

4.5

0.0

90.7

0.0

0.0

0.4

0.4

0.0

nucleoli

(g)

0.0

0.0

1.4

1.4

0.0

0.0

86.5

9.5

1.5

0.0

0.0

nucleus

(h)

0.7

0.0

0.0

0.0

0.0

0.7

2.7

96.0

0.0

0.0

0.0

peroxisomes

(i)

0.0

0.0

0.0

2.8

0.0

2.8

0.0

0.0

88.7

5.6

0.0

plasma membrane

(j)

2.1

0.0

0.0

1.0

0.0

2.1

0.0

0.0

0.0

94.9

0.0

  1. Each row shows the classification results for a specific protein location. Therefore, a single entry denotes the fraction of images from a specific protein location (row), which were associated with a specific class label (column). In the case of a correct classification, the row index and the column index are equal. The column marked by (*) shows the fraction of protein distribution patterns which were considered as unknown.