作者: Ian W. Renner , David I. Warton
DOI: 10.1111/J.1541-0420.2012.01824.X
关键词: Principle of maximum entropy 、 Point process 、 Species distribution 、 Poisson regression 、 Mathematical optimization 、 Poisson point process 、 Unification 、 Lasso (statistics) 、 Equivalence (measure theory) 、 Ecology 、 Mathematics
摘要: Summary Modeling the spatial distribution of a species is fundamental problem in ecology. A number modeling methods have been developed, an extremely popular one being MAXENT, maximum entropy approach. In this article, we show that MAXENT equivalent to Poisson regression model and hence related point process model, differing only intercept term, which scale-dependent MAXENT. We illustrate improvements follow from these relations. particular, approach facilitates for choosing appropriate resolution, assessing adequacy, LASSO penalty parameter, all currently unavailable The equivalence result represents significant step unification literature.