学术文献库

In the past decades, the growing amount of network data has lead to many novel statistical models. In this paper we consider so called geometric networks. Typical examples are road networks or other infrastructure networks. But also the neurons or the blood vessels in a human body can be interpreted as a geometric network embedded in a three-dimensional space. In all these applications a network specific metric rather than the Euclidean metric is usually used, which makes the analyses on network data challenging. We consider network based point processes and our task is to estimate the intensity (or density) of the process which allows to detect high- and low- intensity regions of the underlying stochastic processes. Available routines that tackle this problem are commonly based on kernel smoothing methods. However, kernel based estimation in general exhibits some drawbacks such as suffering from boundary effects and the locality of the smoother. In an Euclidean space, the disadvantages of kernel methods can be overcome by using penalized spline smoothing. We here extend penalized spline smoothing towards smooth intensity estimation on geometric networks and apply the approach to both, simulated and real world data. The results show that penalized spline based intensity estimation is numerically efficient and outperforms kernel based methods. Furthermore, our approach easily allows to incorporate covariates, which allows to respect the network geometry in a regression model framework.