Skip to main content

Table 2 Ranking of the top ten most important features (out of 101) for the selected RF model

From: Modelling and mapping tick dynamics using volunteered observations

Position

Feature

Importance

1

ev-365

15

2

rh-30

11

3

tmax-365

7

4

prec-90

4

5

prec-3

4

6

ev-90

3

7

rh-365

3

8

tmin-365

2

9

prec-365

2

10

tmax-90

2

  1. The sum of the feature importance for all features provided by RF equals to 1, but to ease the interpretation of results we multiplied it by a hundred to have natural numbers. As seen, features involving atmospheric water levels (i.e. evapotranspiration and relative humidity) are found to be important to predict tick activity, since they appear several times in the current ranking
Back to article page

International Journal of Health Geographics

ISSN: 1476-072X