Elad Hazan

@HazanPrinceton

machine learning and optimization @ Google Brain & Princeton

Vrijeme pridruživanja: travanj 2018.

Tweetovi

Blokirali ste korisnika/cu @HazanPrinceton

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

  1. prije 13 sati

    The latest on non-stochastic control from our lab, a new method for control with only partial observation: Joint work with stellar grad students Max Simchowitz (Berkeley) and Karan Singh (Princeton)

    Poništi
  2. proslijedio/la je Tweet

    The 2020 Information Theory and Applications Workshop kicks off on Sunday here in San Diego! Researchers from around the country are gathering to apply theory (ie deep learning, ML) to diverse areas of science and engineering. Schedule and more info:

    , , i još njih 3
    Poništi
  3. 22. sij

    New results on non-stochastic control without knowing the system! joint work with and Karan Singh. You may notice the name is a tribute to one of my favorite papers of all times by Auer, Cesa-Bianchi, Freund and Schapire (a must read!)

    Poništi
  4. 14. pro 2019.

    Tune in to my student Karan Singh's talk on non-stochastic control, selected for best paper award at NeurIPS RL workshop later today:

    Poništi
  5. 9. pro 2019.
    Poništi
  6. 4. pro 2019.

    Postdoc positions available at COS and EE Princeton, in the theoretical machine learning group, pls see application details below by .

    Poništi
  7. 2. pro 2019.

    Interested in collaborating with Google AI Princeton on SOTA research? Please apply here: (specify Princeton) Ph.D. students with strong engineering/coding skills required to help advance our large-scale control library, details:

    Poništi
  8. 22. stu 2019.

    Real pleasure to give a colloquium at lovely EPFL about our groups' work on non-stochastic control. Many excellent questions and control researchers in the audience.

    Poništi
  9. 7. stu 2019.

    I highly recommend this forthcoming talk by about foundational work on boosting and the Frank-Wolfe method!

    Poništi
  10. 31. lis 2019.
    Poništi
  11. 14. lis 2019.

    Proud advisor moment: excellent paper by my former student Dan Garber, an innovator in projection free methods.

    Poništi
  12. proslijedio/la je Tweet
    7. lis 2019.

    Conventional wisdom: "Not enough data? Use classic learners (Random Forests, RBF SVM, ..), not deep nets." New paper: infinitely wide nets beat these and also beat finite nets. Infinite nets train faster than finite nets here (hint: Neural Tangent Kernel)!

    Prikaži ovu nit
    Poništi
  13. 19. ruj 2019.

    Rethinking control: excited to announce work from Princeton and : the first efficient algorithm w. log-regret ("fast rate") for control of linear dynamical systems with changing costs, studied since the works of Kalman:

    Poništi
  14. 29. kol 2019.

    Test drive with my colleague Prof. Shai Shalev Schwartz, and the self-driving car he is pioneering in Mobileye. I'm a believer!

    Poništi
  15. 17. srp 2019.

    Rethinking control: blog post on our latest work @ Princeton and Google AI on boosting for dynamical systems and control: (w. Naman Agarwal, Nataly Brukhim and Zhou Lu)

    Poništi
  16. proslijedio/la je Tweet
    11. srp 2019.

    Remember matrix completion? Deep linear nets solve it better than the old nuclear norm algorithm. Analysis requires going beyond traditional optimization view and understanding . Blog post by Nadav and Wei: . Paper

    Poništi
  17. proslijedio/la je Tweet
    13. lip 2019.

    Speaking tomorrow (Friday) 2pm in workshop on Theor. Physics in Deep Learning. Title: "Is Optimization a sufficient language to understand deep learning?" (Also, grad Orestis speaking Thurs 4pm about our work on word2vec-like methods for representation learning.)

    Prikaži ovu nit
    Poništi
  18. 11. lip 2019.

    finished compiling lecture notes from my course on optimization for machine learning: (comments/suggestions welcome!)

    Poništi
  19. proslijedio/la je Tweet

    Computer science () is getting a new home at thanks to a gift from ‘76 and .

    Prikaži ovu nit
    Poništi
  20. 17. svi 2019.

    Wonderful trend in our department, hope to see the same in other universities

    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:

    ·