Machine Learning and Compression: Exploring the connections
Detalles
At Regus shared office space, Waterloo.
180 Northfield Dr W Unit 4, Waterloo, ON N2L 0C7
Dr. Zhiying (Gin) Jiang of UWaterloo will walk us through some of the math behind learning from and compressing data, using information theory to elucidate, inform, and sometimes outperform transformer based LLMs!
Background reading for this talk: https://arxiv.org/pdf/2206.11573.pdf
Links to Gin's research and projects for background reading: https://bazingagin.github.io/
Hosting kindly sponsored by Autodesk
Machine Learning
Data Science
Mathematics
Computational Linguistics
Information Science