Bekah Overdorf

@bekah_overdorf

Post-doc at EPFL working in Machine Learning and Privacy

Vrijeme pridruživanja: veljača 2013.

Tweetovi

Blokirali ste korisnika/cu @bekah_overdorf

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @bekah_overdorf

  1. proslijedio/la je Tweet
    20. sij

    Some news: The new version of our paper "POTs: Protective Optimization Technologies" with , , and will appear in FAT* 2020 on Jan 27! Paper: What are protective optimization technologies? Thread 👇

    Prikaži ovu nit
    Poništi
  2. proslijedio/la je Tweet
    26. srp 2019.

    . and won the best workshop paper at 2019 - "Subtle Censorship via Adversarial Fakeness in Kyrgyzstan"

    Poništi
  3. 20. srp 2019.

    The wonders of AI. translating my nasal spray. Tbf, it would solve the immediate problem.

    Poništi
  4. proslijedio/la je Tweet
    19. srp 2019.

    The winner of Best HotPETs Talk is Subtle Censorship via Adversarial Fakeness in Kyrgyzstan by Christopher Schwartz and Rebekah Overdorf

    Prikaži ovu nit
    Poništi
  5. proslijedio/la je Tweet

    For everyone who believes that TLS is enough 🤦‍♀️ Kazahkstan starts officially doing MitM attack for all users. Providers sent text messages requesting users to install root certificate and started to intercept traffic.

    Prikaži ovu nit
    Poništi
  6. proslijedio/la je Tweet
    18. srp 2019.
    Prikaži ovu nit
    Poništi
  7. proslijedio/la je Tweet
    18. srp 2019.

    Proud of our colleagues that have won this year's Caspar Bowden PET Award for their outstanding contribution to the deployment of privacy enhancing technology

    Poništi
  8. 16. srp 2019.

    Sounds like the beginning of a . How can we prevent detection of mental illness on social media to be used by, e.g., insurance companies. by and at .

    Poništi
  9. 14. svi 2019.

    I believe theres something similar in the US as well with insurance companies. I’ve heard of putting step trackers on dogs, or better yet, ceiling fans.

    Poništi
  10. 2. svi 2019.
    Poništi
  11. proslijedio/la je Tweet
    27. tra 2019.
    Poništi
  12. proslijedio/la je Tweet
    4. tra 2019.

    Not sure who this (joke) slide will offend more; philosophers, machine learning ppl or functional programmers!

    Poništi
  13. 4. tra 2019.

    Thanks everyone for a great session. Despite the title and the photo of pots and pans in the background, and I will not be able to help you organize your counter space, revolutionary or otherwise.

    Prikaži ovu nit
    Poništi
  14. 3. tra 2019.

    There’s been a lot of great talk at about the harms of tech like predictive policing & facial recognition & how simple “fixes” aren’t enough. Come to our session in The Gallery in a few mins to continue the conversation about how to fight back.

    Prikaži ovu nit
    Poništi
  15. 29. sij 2019.
    Prikaži ovu nit
    Poništi
  16. 29. sij 2019.

    It's my birthday, so come see my 9:30 talk at about POTs and how to counter the negative effects of machine learning and optimization systems.

    Prikaži ovu nit
    Poništi
  17. proslijedio/la je Tweet
    7. pro 2018.

    Don't miss presenting our "Questioning the assumptions behind fairness solutions" at , rooms 511 ABDE, before lunch and after 3pm. See more in the blog post: And even more in the paper:

    Poništi
  18. 6. pro 2018.

    Wanna learn more about what happens when the service provider doesn't have incentives for ? Check out our Poster tomorrow at Critiquing and Correcting Trends in Machine Learning

    Poništi
  19. proslijedio/la je Tweet
    3. pro 2018.

    Speaking of human sides of ML ( tutorial right now), will present our critique of the assumptions behind fairness at the Critiquing and Correcting workshop on Friday. 1/3

    Prikaži ovu nit
    Poništi

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.

    Možda bi vam se svidjelo i ovo:

    ·