Patrones de incidencia del deterioro del arbolado urbano de Bogotá

Abstract

In this paper are presented geospatial relations between Bogota urban trees with the registers of average annual pollution, roads’ influence area and places of interest (Cadastre). For the variables of study there has been used data for year 2007. The methodology used in this case is CRISP-DM (Cross Industry Standard Process for Data Mining). The relations’ patterns have been evaluated under the algorithms of decision trees which define the relevant variables for each Dataset and cluster analysis which analyses the relevant variables regarding the evaluated types of affectations, assuming that the relevant variables of each Dataset were specie, general affectation percentage, trunk affectation percentage, places of interest, roads’ width, number of lanes and District locations. In results, is has been found that species like Urapan and Sauco were those that presented relations with the evaluated variables, being the Urapan the one that related most with the rest of the variables in each Dataset. This research will serve for decisions making regarding the management of urban trees in Bogota.