Two DATACTIVE team members helped coordinate the upcoming Amsterdam Privacy Conference (APC) but withdrew due to concerns over Palantir sponsorship. See statement here: data-activism.net/2018/09/why-we…

Accepted at #AISTATS2019: arxiv.org/abs/1806.00892. We study safe exploration in recommender systems, where the learner is constrained to take actions no worse than a risk-threshold. We answer the question: "what is the cost paid in terms of regret for playing safe?"

How to cluster using bandits? In top-m arm identification, the experimenter has to specify m. What if the algo found the "best" m for you? Check out Maxgap Bandit at #NeurIPS. Joint work wth Ardhendu Tripathy & Rob Nowak. Video:youtube.com/watch?v=fzynRW… Paper:arxiv.org/abs/1906.00547

Congrats Shweta! Proud to have worked with you Oneirix Labs x.com/HamedZamani/st…

#BlackLivesMatter

So proud of and miss my university town!

SPREAD THE WORD: I'M HIRING PHD STUDENTS FOR FALL 2021 AT UPF Barcelona / @DTIC_UPF!! The project is about reinforcement learning theory and is funded by my recent European Research Council (ERC) Starting Grant. (#ERCStG) Details: cs.bme.hu/~gergo/jobs.ht… Deadline: Feb 12, 2021

Excited to organize the "Data-efficient Machine Learning (DeMaL)" workshop The Web Conference along with Chandan Reddy and Nikhil Rao. The call for papers is available at demalworkshop.github.io. Consider submitting your relevant work, see topics below. #www2021

How the tobacco industry beat science, how this is repeating in politics & why facts/truth are overrated. Must read timharford.com/2017/03/the-pr…

#AISTATS2018 was over and it really felt like a fun and warm family gathering. x.com/mpd37/status/9…