Year of Publication: Jul - 2013
Page Numbers: 229-233
Authors: Qun Wu, Xinchao Jiang, Yue Sun, Suoju He
Conference Name: The Third International Conference on Digital Information and Communication Technology and its Applications (DICTAP2013)
- Czech Republic


The purpose of this paper is to find a clear geography distribution of trade areas, with the help of cluster method DBSCAN and Raleigh Law. DBSCAN algorism is a density-based cluster method. In this thesis, DBSCAN is used to define the centrality of Trade Area strictly, ensuring that it is a part of real Trade Area without any doubt. Assume that the central region is signed as ---. Original Raleigh Law is to find the boundary point, so called zero sale/cut off point, which receives the equal forces from different places of high thickness of populations. After deriving it into two dimensional plane, we find that when there are two places, B and C, of different density of populations, all the boundary points between them form a boundary carve shaping a circle. And the place C with less population is closed in the circle. For all points within the circle C has lager trade force than B. So that--- can be seen as a small trade area. Taking advantages of this feature, we can enlarge--- by combining it with several small trade areas so that the goal to reach the boundary of real trade area is achieved.