Skip to main content
. 2018 Apr 17;18:183. doi: 10.1186/s12879-018-3066-0

Fig. 3.

Fig. 3

Variable importance of (a) Random Forest and (b) Gradient Boosting models. Meteorological factors (blue) and its corresponding delayed or lagged meteorological effects (LG; green). Relative contribution of Gradient Boosting models adds up to 100%