The aim of this paper is to group territorial units in areas of high intensity, using SaTScan and Seg-DBSCAN clustering methods to aggregate adjacent spatial units that are homogeneous with respect to the phenomenon being studied. SaTScan scans the region of interest with a moving window and compares a smoothing of the intensity inside and outside it so that units belonging to contiguous windows with similar intensity are aggregated into a cluster. On the other hand, Seg-DBSCAN, a new version of DBSCAN, limits the arbitrariness of the choice of input parameters and identifies clusters as dense regions in space. As an application we analyze geo-referenced data concerning housing problems in Bari and we propose a comparison between the two methods presented.
Comparing SaTScan and DBSCAN methods in spatial phenomena
P. Perchinunno;L'ABBATE S;
2011-01-01
Abstract
The aim of this paper is to group territorial units in areas of high intensity, using SaTScan and Seg-DBSCAN clustering methods to aggregate adjacent spatial units that are homogeneous with respect to the phenomenon being studied. SaTScan scans the region of interest with a moving window and compares a smoothing of the intensity inside and outside it so that units belonging to contiguous windows with similar intensity are aggregated into a cluster. On the other hand, Seg-DBSCAN, a new version of DBSCAN, limits the arbitrariness of the choice of input parameters and identifies clusters as dense regions in space. As an application we analyze geo-referenced data concerning housing problems in Bari and we propose a comparison between the two methods presented.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
