In the final episode of our re:Invent series, we’re joined by Thorsten Joachims, Professor in the Department of Computer Science at Cornell University.
Subscribe: iTunes / Google Play / Spotify / RSS
Thorsten participated at the conference’s AI Summit, presenting his research on “Unbiased Learning from Biased User Feedback.” In our conversation, we take a look at some of the inherent and introduced biases in recommender systems, and the ways to avoid them. We also discuss how inference techniques can be used to make learning algorithms more robust to bias, and how these can be enabled with the correct type of logging policies.
About Thorsten
Mentioned in the Interview
- Presentation: Unbiased Learning from Biased User Feedback
- Presentation: Deep Learning from Logged Interventions
- Sign up for our AI Platforms eBook Series!
- TWIML Presents: AWS re:Invent Series page
- TWIML Online Meetup
- Register for the TWIML Newsletter
“More On That Later” by Lee Rosevere licensed under CC By 4.0
Joseph Catanzarite
Thorsten discusses using counterfactual offline A/B tests to approximate real causal inference. I did not know this was possible, and the talk inspires me to delve deeper into the subject.