Clustering Spatial Data Using a Generalized Linear Mixed Model (GLMM) with Application to Integrated Pest Management

• Activity Number: 659
• Type: Contributed
• Date/Time: Thursday, August 4, 2011 : 10:30 AM to 12:20 PM
• Sponsor: Biometrics Section
• Author(s): Zhanpan Zhang*+ and Daniel R. Jeske and Xinping Cui
• Companies: University of California at Riverside and University of California at Riverside and University of California at Riverside
• Keywords: Clustering ; Spatial GLMM ; Integrated Pest Management

Abstract

It is important to quantify the density of spatially distributed pests when targeting use of pest control methods. However, pest assessment is traditionally done via the hypothesis testing procedure assuming data are identically independent. In this paper we introduce the spatial Generalized Linear Mixed Model (GLMM), and propose a model-based clustering to simultaneously cluster rows and columns of the spatial data. To obtain the optimal clustering, we develop two heuristic algorithms to avoid the extremely high computational complexity. We demonstrate the utility and power of our proposed method through simulation studies and application to an integrated pest management study, in which we combine the spatial clustering with hypothesis testing to make pest assessment more realistic.