Sliding window correlation analysis for dengue-climate variable relationship

• Main Authors: Thiruchelvam, L., Asirvadam, V.S., Dass, S.C., Daud, H., Gill, B.S.
• Format: Conference or Workshop Item
• Published: American Institute of Physics Inc. 2016
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85006001679&doi=10.1063%2f1.4968149&partnerID=40&md5=87fc9013bcd7c8bcbf5e57fef7225dd0; http://eprints.utp.edu.my/30621/
• Institution: Universiti Teknologi Petronas

This study discussed building of sliding windows to analyze the relationship between dengue incidences and weather variables of mean temperature, relative humidity and rainfall, across the timeline. A window sized of 20 was selected and applied to find correlation between dengue incidences and each of the weather variable. A few time lag of zero, two, four, six, and eight is compared and the time lag with best correlation is selected for each weather variable. Study did not found a good insight for analysis using mean temperature and relative humidity. For both these variables, it was suggested dengue incidences is better measured using fluctuation of maximum and minimum values. Analysis using rainfall variable was found to vary across the timeline in magnitude and direction of the correlation. Time lag of eight was found to be the most significant explaining the relationship between dengue incidences and rainfall variable. © 2016 Author(s).