Spatial Model for Cholera Case Prediction

Spatial variation can be seen in many kind of aquatic disease such as Cholera. In Bangladesh Cholera has been an epidemic disease for many years. In this study the areal effect of Cholera spread is checked for a small area of Bangladesh named Matlab for the period of 2005-2009. Different model has been used to check and compare the variation. The exponential correlation model fits better for this data. A geo-statistical model is also fitted for Cholera case prediction. From the analysis we have found that the range of spatial variation is approximately 3 km, which means that the variability of occurrence of Cholera is valid for shorter range. Model parameters are estimated by Mote Carlo maximum likelihood approach. Predictions are done under Trans-Gaussian kriging. Contour plot has been made to show the predicted cholera cases in different location of that area.

Autorentext

Graduated from Institute of Statistical Research and Training, University of Dhaka in 2011. Currently doing PhD at University of Calgary under the supervision of Dr. Jingjing Wu. My research interest is application of statistics in biological science.