Tree-based classifiers and GIS for biological risk forecasting
Cesare Furlanello, Stefano Merler, C. Chemini · 1997
A tree-based classifier of tick presence, developed to estimate the risk of tick-borne diseases, is currently being interfaced and applied with a Geographical Information System (GIS) based on GRASS (USA-CERL Geographic Resources Analysis Support System). Environmental factors (altitude, substratum, vegetation, exposition, etc.) and tick sampling are used to predict occurrence of the parasite Ixodes ricinus on a target territory. A Tcl/Tk interface to GRASS has been developed for regional data from the Sistema Informativo Ambiente Territorio (S.I.A.T.) of the Autonomous Province of Trento. The system predictions are consistent with the current knowledge of tick ecology and can be used for an effective control of tick-borne diseases. 1 Introduction This paper describes a computational environment for predictive classification models based on georeferenced data. The system is the result of a joint research initiative in computational methods for biological risk assessment in ecology by ...