May 2020

Fri
May 01
Special Event

The Engineering Scholarship Application deadline has been extended to May 1st this year.  

 

Fri
May 08
https://utexas.zoom.us/j/98653405801

In this talk, we aim to quantify the robustness of distributed training against worst-case failures and adversarial nodes. We show that there is a gap between robustness guarantees, depending on whether adversarial nodes have full control of the hardware, the training data, or both. Using ideas from robust statistics and coding theory we establish robust and scalable training methods for centralized, parameter server systems.