Spatiotemporal Kernel Density Estimation

When we conduct a study on a set of geographic events or objects, we usually need to obtain the distribution characteristics of the events (or objects) in the study region so that we can analyze the target events as a whole and see if there are underlying factors influencing their spatial distribution. There are different ways to obtain the distribution characteristics of the study objects, either by direct analysis of the distribution of the target events through mathematical tools such as histograms or by quantitative study of the attribute values of the target events using mathematical formulas. Among the existing research approaches, kernel density estimation (KDE) is highly valued by researchers as evident by its increasing use. KDE is a method to study the characteristics of data distribution from the data itself.