INVESTIGATING POPULATION BALANCE IN THE STATE OF GEORGIA USING SPATIAL CLUSTERING

Abstract

Identification of population ratio disruption in the population structure is one of the challenges that every country faces. Population aging is a kind of demographic abnormality that lack of attention causes population problems. A timely warning about the aging of society can be useful for planning about having children on the one hand and providing suitable facilities for the elderly on the other hand. For example, the capital of Japan is a good example of an urban environment suitable for the elderly.One of the anomaly detection tools is spatial clustering using scan statistics. In the last three decades, the scan statistic method has been a very important and active field in statistical research. Identifying areas on geographic maps, where the concentration of points (elderly, sick, criminals, certain animal species, etc.) is significant, is important in many fields such as epidemiology, politics, criminology, zoology, and so on. With the help of scan statistic method, spatial clusters can be identified.In this article, we introduce the scan statistic method based on Poisson distribution. Using simulation, we investigate the efficiency of this method in identifying spatial clusters. Based on the results obtained from the simulation, the Poisson scan statistic method is a suitable method for detecting anomalies in the count spatial data. As an application of spatial clustering, we consider the population structure of the state of Georgia and identify areas where the elderly population is significantly high. These areas should be prioritized in the implementation of population reform programs.

Keywords

• spatial clustering
• scan statistic
• monte-carlo hypothesis testing
• poisson distribution
• likelihood ratio