Csaba Szepesvari

@CsabaSzepesvari

"If there is not folly in the world, then the world itself is folly. You must understand that mistakes are not always regrets." - Paul Tobin, Bandette🤠

Vrijeme pridruživanja: veljača 2016.

Tweetovi

Blokirali ste korisnika/cu @CsabaSzepesvari

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

  1. 12. sij
    Prikaži ovu nit
    Poništi
  2. 12. sij

    Cool universality argument for SGD with FF neuralnets: Take any learning alg A for learning Boolean functions without noise from a sample of size n. Then there is a NN architecture G(A,n) such that SGD+G(A,n)+Any reasonable loss with sequential processing "implements" A.

    Prikaži ovu nit
    Poništi
  3. 9. sij
    Poništi
  4. 7. sij

    Association for Computing Machinery: ACM Support Open Access - Sign the Petition! via

    Poništi
  5. proslijedio/la je Tweet
    6. sij

    My great pleasure to share the CFP for the MLJ Special Issue on RL for Real Life (again): (With Alborz Geramifard, , , Tao Wang). Deadline: March 5, 2020.

    Poništi
  6. proslijedio/la je Tweet
    29. pro 2019.
    Poništi
  7. 17. pro 2019.

    This seems to be a good indication that our choices are governed by sophisticated pattern matching rather than symbolic reasoning, though clearly people can also use symbolic reasoning when it is really needed. Fascinating.

    Poništi
  8. 11. pro 2019.

    Only 4 days left till the end of the application deadline for the Simons Fellowship. I hope that many strong junior researchers (0<=x<=6, x=years after PhD) interested in RL will apply. Recall that our RL theory program will start in the fall of 2020

    Poništi
  9. 6. pro 2019.

    Pre-neu_rips exploration (can you spot the 5th player?)

    Poništi
  10. proslijedio/la je Tweet
    6. pro 2019.

    hunting for wild likes & retweets in the woods of Canada---where are they hiding??

    Poništi
  11. 30. stu 2019.
    Poništi
  12. 30. stu 2019.

    I really liked this episode of : Talking sanity in this world.

    Poništi
  13. proslijedio/la je Tweet
    27. stu 2019.

    If you are interested in graduate studies on reinforcement learning, consider applying to the Department of Computer Science, University of Toronto and work with me.

    Prikaži ovu nit
    Poništi
  14. 25. stu 2019.

    "The world has always been a sorry and confused sort of place-yet poets and artists and scientists have ignored the factors that would, if attended to, paralyze them." -- good to remember this at times. When was this written and by whom?

    Poništi
  15. 17. stu 2019.

    To the attention of friends of : After all those years, finally, our home, from is live on twitter.

    Poništi
  16. proslijedio/la je Tweet
    4. stu 2019.

    Folks, have a winner. For the stochastic bandit, this gem from Odalric-Ambrym Maillard's thesis, called EwS, works incredibly well and gives explicit probabilities on the arms. I compare EwS with Thompson & Tsallis in

    Prikaži ovu nit
    Poništi
  17. proslijedio/la je Tweet
    6. stu 2019.

    Post-Doc position to work with me on Online Learning/Optimization/Statistical Learning Theory problems. Ideal candidate has multiple theory papers at NeurIPS/ICML/COLT and hates parameters in ML 😀 E-mail me to apply, we can also meet at NeurIPS. Please RT & share it!

    Poništi
  18. 30. lis 2019.

    Congrats , well deserved! On to matching constants in upper and lower bounds and beyond!

    Poništi
  19. 30. lis 2019.

    To the attention of strong final year PhD students, junior faculty in CS/Theory/..! Excellent opportunity to stay at Berkeley while the 'Theory of RL' and other programs are happening. Please pass it along to relevant candidates.

    Poništi
  20. 11. lis 2019.

    I'm not shocked (my nerves are like a steel rope), but this is still important.

    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:

    ·