Deep Reinforcement Learning for Conservation Decisions

TL;DR

This paper explores how reinforcement learning can improve conservation decision-making by handling dynamic environments, requiring less data, and integrating with existing ecological models, with practical examples and code resources.

Contribution

It introduces reinforcement learning as a novel approach for conservation, demonstrating its applicability and benefits in ecological management problems.

Findings

RL effectively models dynamic conservation environments

RL requires less data than traditional methods

Practical code examples facilitate adoption by researchers

Abstract

Can machine learning help us make better decisions about a changing planet? In this paper, we illustrate and discuss the potential of a promising corner of machine learning known as _reinforcement learning_ (RL) to help tackle the most challenging conservation decision problems. RL is uniquely well suited to conservation and global change challenges for three reasons: (1) RL explicitly focuses on designing an agent who _interacts_ with an environment which is dynamic and uncertain, (2) RL approaches do not require massive amounts of data, (3) RL approaches would utilize rather than replace existing models, simulations, and the knowledge they contain. We provide a conceptual and technical introduction to RL and its relevance to ecological and conservation challenges, including examples of a problem in setting fisheries quotas and in managing ecological tipping points. Four appendices with annotated code provide a tangible introduction to researchers looking to adopt, evaluate, or extend these approaches.