Table 1 Operation errors of the proposed structure when using different numbers of cascade levels.

From: An interpretable ensemble structure with a non-iterative training algorithm to improve the predictive accuracy of healthcare data analysis

Cascade levels

Maximum residual error

Mean absolute error

Mean squared error

Median absolute error

Training time, seconds

Training mode

1

1.808

0.143

0.046

0.096

62.1

2

1.361

0.056

0.008

0.036

87.5

3

0.948

0.052

0.007

0.034

100.0

4

0.969

0.050

0.006

0.033

114.8

Test mode

1

2.163

0.146

0.049

0.097

–

2

1.514

0.056

0.008

0.036

–

3

1.400

0.053

0.007

0.034

–

4

1073.098

0.071

15.511

0.034

–

Back to article page