1  Introduction

This is a blog post/discussion about using Nonhomogeneous Poisson Point Processes for modelling Presence-Only data. This topic became a central unit of my PhD research, and so I wanted to condense the most important information about this data/model in a nice quarto document. Then I realized I could just make this into a blog post, so that is why it is here, now. In this document I discuss models for presence-only data, namely the Thinned Nonhomogeneous Poisson Process (NHPP) and also, briefly, the Log Gaussian Cox Process (LGCP). The initial planned layout is as follows: Introduce presence-only data, Poisson Process models and the data generating process. Then, proceed to fit a Thinned Nonhomogeneous Poisson Process in the Bayesian programming language Stan on simulated presence-only data. Finally, fit a LGCP for comparison.

The rest of this document follows the above description. Please reach out to max.savery@ugent.be with any questions.