Skip to main content

Table 3 Mean decrease in accuracy of the established random forest model if removing the variable.

From: Infectious bursal disease virus: predicting viral pathotype using machine learning models focused on early changes in total blood cell counts

 

Classical pathotype (Cla strain)

Highly attenuated pathotype (i vaccine strain)

Attenuated pathotype (i + vaccine strain)

Strictly immune-suppressive pathotype (im1 and im2 strains)

Very virulent pathotype (Vv1 and Vv2 strains)

Clinical score at 4 dpi

0.02

0.06

0.05

0.07

0.03

Log (r)

0.03

0.03

0.03

0.03

0.00

Log (thr)

−0.01

0.03

0.08

0.04

0.10

Log (m)

0.06

0.00

0.00

0.00

0.04

Log (g)

0.06

0.07

0.15

0.02

0.04

Log (b)

0.02

0.45

0.12

0.04

0.03

Log (t)

0.06

0.08

0.01

0.44

0.17

Log (n)

0.01

0.25

0.01

0.04

0.01

Log (u)

−0.02

−0.01

0.13

0.01

0.01

  1. Features analysis was performed using the RandomForestExplainer [44].
  2. r : erythrocytes blood concentration, thr:  thrombocytes blood concentration, m: monocytes blood concentration, g: granulocytes blood concentrations, b: lymphocytes B blood concentration, t: lymphocytes T blood concentration, n: bursal viral load, u: uric acid blood concentration at 4 dpi.