A Fuzzy C-Means-Based Algorithm for the Surveillance of Dengue Cases Distribution in Local Communities

Abstract

The analysis of disease occurrence over the smallest unit of a region is critical in designing data-driven and targeted intervention plans to reduce health impacts in the population and prevent spread of disease. This study aims to characterize groups of local communities that exhibit the same temporal patterns in dengue occurrence using the Fuzzy C-means (FCM) algorithm for clustering spatiotemporal data and investigate its performance in clustering data on dengue cases aggregated yearly, monthly and weekly. In particular, this study investigates similar patterns of Dengue cases in 129 barangays of Baguio City, Philippines recorded over a period of 9 years. Results have shown that the FCM has promising results in grouping together time series data of barangays when using data that is aggregated weekly.