Table 1 R-squares of CNNRNN-Res, CNNRNN and RNN predictions on test set in different age groups and horizons.

From: Using deep learning to predict the hand-foot-and-mouth disease of enterovirus A71 subtype in Beijing from 2011 to 2018

Model

Age group

Horizon (months)

1

2

4

6

8

10

12

CNNRNN-Res

0–3

0.8041

0.7239

0.3660

0.6908

0.4632

0.8492

0.9032

3–6

0.9174

0.8908

0.4038

0.7711

0.6889

0.8512

0.8218

 > 6

0.8823

0.8651

0.3447

0.7185

0.7779

0.7730

0.7057

Total

0.9062

0.8467

0.4647

0.7654

0.6274

0.8417

0.8538

CNNRNN

0–3

0.8819

0.5745

0.3596

0.3093

0.3734

0.8075

0.8715

3–6

0.8769

0.6853

0.5535

0.4365

0.3966

0.7682

0.8310

 > 6

0.8306

0.6543

0.5927

0.3844

0.4263

0.6798

0.7523

Total

0.8845

0.6659

0.5110

0.4244

0.4682

0.8106

0.8661

RNN

0–3

0.8662

0.4861

0.4919

0.6955

0.7738

0.8208

0.8291

3–6

0.8881

0.6684

0.6101

0.7049

0.6855

0.8250

0.7032

 > 6

0.8615

0.6754

0.6190

0.6969

0.5862

0.7849

0.6464

Total

0.8911

0.6302

0.5913

0.7255

0.7292

0.8442

0.7722

  1. The number in bold indicates the maximum value in a certain horizon and age group. ‘Total’ means the total number of cases of EV-A71 subtype in Beijing.
Back to article page