UCL School of Management is delighted to welcome Professor Jue Wang, Queens University, to host a research seminar discussing ‘Analytics Going Wild: Learning and Management of Wildlife’.

Abstract

Ecosystems contribute to over half of the global GDP, but it is approaching a dangerous tipping point; wildlife populations have declined by 73% in the last 50 years, and half of the species are heading toward extinction in the next 100 years. A major challenge in biodiversity conservation is the scarcity of funding and information. As species populations decline, they become increasingly difficult to find. Data is costly and time-consuming to collect, and high-stake decisions must be made based on small, low-quality data. Further, the species population can deteriorate rapidly, leading to extinction before sufficient data can be gathered. Therefore, how to allocate limited resources in a noisy, dynamic, and uncertain environment is a pressing challenge. The scale, uncertainty, and spatial structure inherent in conservation demand innovations in decision science. In this talk, we use two case studies to illustrate the cross-fertilization between biodiversity conservation and partially observable Markov decision processes (POMDP). We will talk about the search and protection of Hainan gibbon, the rarest primate on earth, in the Hainan Tropical Rainforest National Park. We also explore the surveillance and management of a devastating wildlife disease, informally known as “zombie deer disease”, across North America. These projects are not possible without interdisciplinary, multi-institutional collaboration with conservation biologists, disease ecologists, and wildlife agency managers.