Using Markov decision processes to find optimal ways to manage ecological systems – Tracy Rout


Managers of ecological systems must make decisions in the face of considerable uncertainty. This arises for three reasons: natural systems are variable and unpredictable, we cannot observe them precisely, and we usually don’t understand exactly how they work. Markov decision processes (MDPs) are a useful tool for formulating ecological management problems and solving them to find the best management strategies under uncertainty. In this lecture, I will show how MDPs are applied to ecological management problems, illustrating this with examples from my own work. I will cover forms of MDPs that can account for different types of uncertainty, including formulations of adaptive management problems.

Download

Comments are closed.