Hello everyone,

Happy New Year! Our first seminar of the semester will be next Wednesday (2024/01/17) at noon in BBB. Yutong Wang will be presenting his research in BBB room 3901. Please fill out the food form before attending, so we can buy enough pizza for everyone.

If you have research to share, please volunteer to present using this link. We already have a presenter for the second January meeting, but we do not have anyone signed up to present starting in February. As a token of gratitude, presenters get to choose a customized meal from a selection of local restaurants, as listed here.

All seminar info is available on the SPEECS website, and a Google calendar link with dates/times/presenters is can be found here. If you have any questions, you can contact Matt Raymond or me directly, or email speecs.seminar-requests@umich.edu. Suggestions are always welcome :)

Speaker: Yutong Wang

Topic: Reinventing the Foundations of Multiclass Classification: Practical Theory for Scalable Algorithms

Abstract: The notion of margin loss has been central to the development and analysis of algorithms for binary classification. To date, however, there remains no consensus as to the analogue of the margin loss for multiclass classification. In this work, we show that a broad range of multiclass loss functions, including many popular ones, can be expressed in the relative margin form, a generalization of the margin form of binary losses. The relative margin form is broadly useful for understanding and analyzing multiclass losses as shown by our prior work (Wang and Scott, 2020, 2021). To further demonstrate the utility of this way of expressing multiclass losses, we use it to extend the seminal result of Bartlett et al. (2006) on classification-calibration of binary margin losses to multiclass. We then analyze the class of Fenchel-Young losses, and expand the set of these losses that are known to be classification-calibrated.

Supplementary link: https://arxiv.org/abs/2311.17778

Mirror: https://websites.umich.edu/~speecsseminar/presentations/20240117/

Thanks,

Rachel Newton