Temporal Network Kernel Density Estimation

Kernel density estimation (KDE) is a widely used method in geography to study concentration of point pattern data. Geographical networks are 1.5 dimensional spaces with specific characteristics, analyzing events occurring on (accidents roads, leakages pipes, species along rivers, etc.). In the last decade, they required extension spatial KDE. Several versions Network KDE (NKDE) have been proposed, each their particular advantages and disadvantages, now regular basis. However, scant attention has given temporal NKDE (TNKDE). practice, when studied happen at time points constrained network, methodologies by geographers tend overlook either network or dimension. Here we propose TNKDE based recent development product kernels. We also adapt classical methods (Diggle's correction, Abramson's adaptive bandwidth selection leave-one-out maximum likelihood). illustrate Montreal road crashes involving pedestrian between 2016 2019.