[Math_club] Math Club Meeting (2/25 @ 5pm): UO Math Alum Talk

[Christina Dorofeev]

Hi everyone,

Math club will hold its last meeting of the term this Wednesday, February 25, at 5pm in University 213. We're excited to welcome Natalie Weaver, a UO math alum, who will be giving a talk titled "Explainable Boosting Machines," followed by a Q&A and a discussion of her professional experience.

Talk Abstract:
Data science is all about trade-offs: bias vs variance, data quality vs quantity, computational cost vs performance, and so on. Until recently, one of the major trade-offs to consider was accuracy vs interpretability: simple models such as linear regressions or individual decision trees tend to have high interpretability but low accuracy, while complex machine learning methods such as extreme gradient boosting and neural networks tend to have high accuracy but low interpretability.

Explainable boosting machines are a modern machine learning approach developed by Microsoft Research that combine the interpretability of a linear regression with the predictive accuracy of a gradient boosted model, effectively dissolving what was once considered a fundamental trade-off in the field of data science.

As always, pizza will be provided. We hope to see you there!

Best wishes,
UO Math Club
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.uoregon.edu/pipermail/math_club/attachments/20260224/302e3443/attachment.html>