Search for: All records

Minimax-fair machine learning minimizes the error for the worst-off group. However, empirical evidence suggests that when sophisticated models are trained with standard empirical risk minimization (ERM), they often have the same performance on the worst-off group as a minimax-trained model. Our work makes this counter-intuitive observation concrete. We prove that if the hypothesis class is sufficiently expressive and the group information is recoverable from the features, ERM and minimax-fairness learning formulations indeed have the same performance on the worst-off group. We provide additional empirical evidence of how this observation holds on a wide range of datasets and hypothesis classes. Since ERM is fundamentally easier than minimax optimization, our findings have implications on the practice of fair machine learning. 

In cooperation with elders of the Northern Arapaho Language and Culture Commission (NALCC), a language revitalization project using virtual reality is being developed, supported by a National Science Foundation grant. The origins of the project are explored, underlying methodologies examined, as well as the important role that the elders of the Northern Arapaho Language and Culture Commission play in guiding the goals of the grant: (a) exploring the potentials of virtual reality in language revitalization; (b) documenting spoken Arapaho language with an emphasis on hunting and animal migration stories and songs related to place names on the Wind River Indian Reservation and other locations in Wyoming and Colorado; and (c) developing virtual reality curricula units for Wind River Indian reservation K–12 schools. Difficulties in conducting research during the covid19 pandemic, especially with Indigenous communities that have been hit hard by the virus, impacted our methodology and project process. This project seeks to provide a blueprint for other scholars interested in working with tribes and grant agencies in using VR in language revitalization. The project engages the questions if and how VR and subsequent technologies can be used as decolonial tools to help reverse language loss and promote culture. 

Minimax-fair machine learning minimizes the error for the worst-off group. However, empirical evidence suggests that when sophisticated models are trained with standard empirical risk minimization (ERM), they often have the same performance on the worst-off group as a minimax-trained model. Our work makes this counter-intuitive observation concrete. We prove that if the hypothesis class is sufficiently expressive and the group information is recoverable from the features, ERM and minimax-fairness learning formulations indeed have the same performance on the worst-off group. We provide additional empirical evidence of how this observation holds on a wide range of datasets and hypothesis classes. Since ERM is fundamentally easier than minimax optimization, our findings have implications on the practice of fair machine learning. 

Abstract Anthropogenic warming has led to an unprecedented year-round reduction in Arctic sea ice extent. This has far-reaching consequences for indigenous and local communities, polar ecosystems, and global climate, motivating the need for accurate seasonal sea ice forecasts. While physics-based dynamical models can successfully forecast sea ice concentration several weeks ahead, they struggle to outperform simple statistical benchmarks at longer lead times. We present a probabilistic, deep learning sea ice forecasting system, IceNet. The system has been trained on climate simulations and observational data to forecast the next 6 months of monthly-averaged sea ice concentration maps. We show that IceNet advances the range of accurate sea ice forecasts, outperforming a state-of-the-art dynamical model in seasonal forecasts of summer sea ice, particularly for extreme sea ice events. This step-change in sea ice forecasting ability brings us closer to conservation tools that mitigate risks associated with rapid sea ice loss.