Torus Talks: Robust Sub-Sampling for Big Data Regression by Jordan Slessor

October 21, 2021

These are the Math and Stats Department Torus Talks — they are meant to showcase the beauty of mathematics in a low-key and fun way.

Big data is a field of study pertinent to statistics and computer sciences alike. Alongside big data's size, speed, and complexity comes several new challenges to our current statistical and computational methods.

Leveraging for big data is a subsampling-based method to approach challenges introduced by the size of big data, and M-estimators' attempt to tackle specifically the increased probability of finding outliers within datasets from Big Data.

Through empirical analysis of contaminated and non-contaminated datasets, the leveraging estimators were found to be less robust when compared to regression M-estimators.

All in the university community are invited. Torus Talks are pitched to a general university audience; all students, staff and faculty are welcome.

This event will be presented online using Zoom. Presenter: Jordan Slessor


Launch meeting


Department of Mathematics and Statistics

Event Details

Date October 21, 2021
Start Time 1:00 p.m.
End Time 2:00 p.m.
Location Online via Zoom
Event Type Lecture/speaker/presentation
Contact Person Karen Buro
Contact Phone 780-633-3911
Contact Email
Event Website Visit Site