Table 3 Importance ranking of predictors for low-path AI model based on Treenet algorithm (SPM).

From: Data mining and model-predicting a global disease reservoir for low-pathogenic Avian Influenza (AI) in the wider pacific rim using big data sets

Predictor rank of importance

Name of predictor

Percent importance ranking

1

Host species*

100

2

Koeppen geiger classification

24

3

Proximity to known poultry farms

13

4

Mean precipitation June

12

5

Proximity to roads

12

6

GLC2000 landcover

11

7

Slope in degrees

11

8

Proximity to coast

10

9

Proximity to pig farms

9

10

Mean precipitation September

8

11

Mean temperature December

8

12

Altitude

7

13

Nationalpark

7

14

Mean temperature March

6

15

Mean precipitation March

6

16

Mean temperature September

5

17

Mean temperature June

5

18

Mean precipitation December

5

19

Located in cyclone area

3

  1. * An attribute that is associated with AI lab data; it was used for the data mining (not landscape prediction surface as such information is not really available on a landscape scale; see also10,15)
Back to article page