Skip to main content

Table 4 Dimension reduction of different accumulative contribution rates of five types of ECG

From: A particle swarm optimization improved BP neural network intelligent model for electrocardiogram classification

Accumulative contribution

Types

The original feature dimension

feature dimension after PCA

Training data

Test data

\(a > 85\%\)

Vpb

15

3

3

Rbbb

15

4

3

Nb

15

3

4

Apb

15

4

3

Lbbb

15

3

4

\(a > 99\%\)

Vpb

15

6

6

Rbbb

15

7

6

Nb

15

6

6

Apb

15

7

6

Lbbb

15

6

7

  1. \(a\) means the selected principal components can represent \(a \times 100\%\) of the original information