Computer Science 691SU - S-Computational Sustainability

Ecosystems across the globe under threat. Defining good conservation policies is difficult because many aspects of ecosystems and our impacts on them are poorly understood. However, new data resources are emerging that can help us understand and manage ecosystems more effecitively, if we can develop the algorithms to understand and use this data well. For example, some algorithmic problems in ecology include: where to place sensors, how to fit models of complex ecological such as bird migration from noisy and incomplete data, and where to place wildlife reserves or corridors to maximize the benefits to threatened species. In this seminar, students will read and discuss recent papers at the intersection of CS and ecology, with a focus on novel work in machine learning, AI, and discrete optimization. The instructor will also present open problems, data resources, and research ideas.

Open to CMPSCI graduate students only. CMPSCI 611 or 688 or 689 LECT 01=3 credits; Lect 02=1 credit. STUDENTS NEEDING SPECIAL PERMISSION MUST REQUEST OVERRIDES VIA THE ON-LINE FORM: https://www.cs.umass.edu/overrides.

Students taking the course for three credits will complete a project or extended survey. Students should have completed a Ph.D.-level course in machine learning or algorithms or receive permission from the instructor.