Want to help solve #ClimateChange? Here’s what #MachineLearning can do. Detailed recommendations for researchers, entrepreneurs, and public/private stakeholders. #ClimateChangeAI
Full paper: arxiv.org/abs/1906.05433
David Rolnick
467 posts
David Rolnick
@david_rolnick
Assistant Professor in Computer Science, @mcgillu / @Mila_Quebec. Co-Founder and Chair, @ClimateChangeAI. MIT @techreview Innovator Under 35. he/him/his
Joined January 2015
- Excited to announce I am starting this fall as Assistant Professor of Computer Science at @mcgillu /@MILAMontreal. I'll be taking grad students and postdocs both in deep learning theory and in machine learning applied to climate change - please get in touch if interested!
- My team is currently looking for (paid) research interns in the following areas of AI-for-climate: catalyst design (comp chem background preferred), climate model emulation (ML or climate sci background), insect ID (computer vision or ecology background). Email me if interested.
- We are looking for new PhD / MSc students! Machine learning innovations for tackling climate change - e.g. remote sensing, climate informatics, materials discovery, & biodiversity monitoring. Apply by submitting applications to *both* @Mila_Quebec (by Dec 1) & @mcgillu (Dec 15).
- What drives innovation in machine learning? In a new paper, we argue that application-driven work is systemically under-valued in the machine learning community, but that it's essential for both innovation and impact. arxiv.org/abs/2403.17381 1/3
- In new work with @KordingLab, we reconstruct the architecture, weights, and biases of unknown deep ReLU networks by observing their output. ReLU nets are piecewise linear functions - a lot of info is in the boundaries between linear regions. Full paper: arxiv.org/abs/1910.00744
- ICLR reviewer: Thanks, you've addressed all my concerns and the paper's great! Same ICLR reviewer: I'm still going to keep my score at Borderline Accept.
- I'm honored to be named to the @techreview list of “Innovators Under 35”, and am thrilled to see increasing interest across the AI community in cross-disciplinary partnerships for climate action. 1/3
- According to ImageNet, fish are dead, birds are American, and 98% of ferrets are wrong. In new work with @SashaMTL (arxiv.org/abs/2208.11695), we look at ImageNet-1k animal photos, which are a quarter of the whole dataset. Working with ecologists, we find major errors & biases. 1/
- Honored to present on AI and climate change to Germany’s Chancellor @OlafScholz and Vice Chancellor Habeck during their visit to @Mila_Quebec. Very happy to see this interest in our group’s research and our policy reports with @ClimateChangeAI.
- We’re also looking for @MilaQuebec interns for: (1) machine learning w/ time series to optimize HVAC energy use (summer or fall), (2) physics-informed ML for climate models (fall), (3) generative models for catalysts using density functional theory (fall). Email me if interested.
- Thrilled to announce our new Global Center in AI and Biodiversity Change (ABC) as one of seven NSF-funded Global Centers in Climate Change and Clean Energy. Prospective postdocs / interns interested in AI + biodiversity, please reach out to me by email. mcgill.ca/research/chann…
- Hi from the rainforest in Panama! For the next week, we will be here testing our AI-enabled camera traps for automated monitoring of insects. Stay tuned for updates on machine learning, moths, and other cool things from Barro Colorado! #AutomatedMonitoringOfInsects
- The space of ReLU networks has well-known symmetries that change the params while preserving the function - but there are other hidden symmetries. Our recent paper with @eligrigsby & Kathryn Lindsey #ICML2023 explores where they hide & how often they occur
