Density-based clustering for bivariate-flow data

Geographical flows reflect the movements, spatial interactions or connections among locations and are generally abstracted as origin-destination (OD) flows. In this context, clustering is a pattern describing group of with adjacent O D points. For data composed two types (bivariate-flow data), bivariate-flow cluster comprising flows, at least one which exhibits pattern. cluster, varying flow density combinations imply different meanings. instance, high-density travel on both weekdays (type A) weekends B) may be associated entertainment, whereas sparse reveal work-related travel. However, identifying clusters still an unsolved problem. To end, we extend bivariate-point method propose density-based for bivariate The simulation experiments verify model robustness. case study, apply to extract Beijing taxi OD periods, identify work-related, tourism, egress return travels. These results demonstrate capability our in detecting clusters.