Censored Distance-Based Intensity Estimation of Spatial Point Processes
研究了利用删失距离估计空间点过程强度的方法，通过最大似然估计和稳健复合估计处理采样重叠和边缘效应，并通过模拟比较了删失与未删失测量的表现。
The use of censoring in conjunction with distance-based estimation of the intensity of a spatial point process is considered. Censoring, in this context, may be motivated by practical considerations such as reduction of sampling effort, but it may also be considered as a method for preventing the overlap of search regions and for eliminating edge effects. Maximum likelihood estimators that utilize censored point-plant measurements or censored T-square measurements are obtained for the case of a homogeneous Poisson process, and some robust ‘compound’ estimators that utilize censored T-square measurements are proposed for more general use. The effect of sampling intensity on the performance of these estimators and their counterparts that utilize uncensored measurements is studied via simulation, with particular attention given to discovering under what circumstances it is advantageous to use censoring to deal with overlap and edge effects.