Kimia Hamidieh (@kimiahmdh) 's Twitter Profile
Kimia Hamidieh

@kimiahmdh

PhD student at @MIT_CSAIL, previously @UofT/@VectorInst

ID: 1311271453980065794

calendar_today30-09-2020 11:47:43

8 Tweet

269 Followers

233 Following

Aleksander Madry (@aleks_madry) 's Twitter Profile Photo
Aleksander Madry
@aleks_madry

a year ago

In ML, we train on biased (huge) datasets ➡️ models encode spurious corrs and fail on minority groups. Can we scalably remove "bad" data? w/ Saachi Jain Kimia Hamidieh Kristian Georgiev Andrew Ilyas Marzyeh we propose D3M, a method for exactly this: gradientscience.org/d3m/

In ML, we train on biased (huge) datasets ➡️ models encode spurious corrs and fail on minority groups. Can we scalably remove "bad" data? w/ <a href="/saachi_jain_/">Saachi Jain</a> <a href="/kimiahmdh/">Kimia Hamidieh</a> <a href="/kris_georgiev1/">Kristian Georgiev</a> <a href="/andrew_ilyas/">Andrew Ilyas</a> <a href="/MarzyehGhassemi/">Marzyeh</a> we propose D3M, a method for exactly this: gradientscience.org/d3m/