A Self-Adjusting Approach to Identify Hotspots

Abstract

Hotspot identification or detection has been widely used in many fields; however the traditional grid-based approaches may incur some problems when dealing with point database. This article expands on three types of mismatch problems in grid-based approach and suggests a point-based approach may be more suitable. Inspired by the DBSCAN algorithm, a self-adjusting approach is then proposed for hotspot detection which overcomes the weakness of parameter sensitivity shared by most clustering approaches. Finally, the data of commercial points of interest of a city is used for demonstration.