Embracing Uncertainty by Rebekah Lindsay & Amina Naidjate '25

Embracing Uncertainty — The Principle of Maximum Entropy for Species Distribution Modeling by Rebekah Lindsay & Amina Naidjate ’25, Wednesday February 12, 1:00 – 1:50pm, North Science Building 015, Wachenheim, Statistics Colloquium

Combined: What happens when a frequentist statistician has incomplete information on the system that they are working with? While Bayesians rely on prior knowledge to fill in the gaps, frequentists adopt alternative methods like maximum entropy models. Under the Principle of Maximum Entropy, such models select probability distributions that maximize uncertainty while satisfying pre-existing constraints. In a world of limited data, they provide a robust framework for estimating distributions using only known information, ensuring no unwarranted assumptions are made. With examples adapted from the work of Paul Penfield, we’ll be exploring the statistical underpinnings of maximum entropy for increasingly complex systems, including an exercise on how it can be utilized in species distribution modeling for ecologists.