Aditya Parameswaran

@adityagp

Assistant Professor | Analytics of, by, and for the People | Past: Stanford PhD, MIT Postdoc, UIUC Prof

Vrijeme pridruživanja: travanj 2008.

Tweetovi

Blokirali ste korisnika/cu @adityagp

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

  1. Prikvačeni tweet
    18. lis 2017.

    My post on the five BIG, EXCITING challenges in enabling data science for "the 99%"

    Prikaži ovu nit
    Poništi
  2. prije 23 sata

    Please consider the environment before training this model.

    Poništi
  3. proslijedio/la je Tweet
    31. sij

    Encourage your students to submit their proposals to this year's VLDB PhD workshop. We have an excellent PC and as keynote speaker. Organized by and me

    Poništi
  4. proslijedio/la je Tweet
    31. sij
    Poništi
  5. 31. sij

    Examples of tools include but not limited to .

    Prikaži ovu nit
    Poništi
  6. 31. sij

    We're seeking interviewees for a study on the use of automated machine learning () tools. The interview will take ~one hour, and you will be compensated for your insights. Please sign up via if you're interested in participating.

    Prikaži ovu nit
    Poništi
  7. proslijedio/la je Tweet

    "How diverse are our conferences? How diverse is our community?" director Magda Balazinska and scientists Carlo Curino and Avrilia Floratou report their experience from organizing SOCC'19:

    Poništi
  8. proslijedio/la je Tweet
    28. sij

    1/ A year ago, we wrote about the potential of serverless, but we felt existing infra had key shortcomings. We spent the last year solving some of the core challenges around FaaS + state management, and we're excited to share our progress:

    Prikaži ovu nit
    Poništi
  9. 28. sij

    Congrats !! Doris is one of 36 fellows from over 1800 applicants.. and one of three awardees. Doris's research is on intelligent tools for exploratory data analysis and data science. Very well deserved!

    Poništi
  10. proslijedio/la je Tweet
    21. sij
    Odgovor korisnicima i sljedećem broju korisnika:

    Brilliant indeed! Also, the tweet is out of date. We hired that year as well. And in DB-related theory we hired and Barna Saha. See , and

    Poništi
  11. proslijedio/la je Tweet
    Odgovor korisnicima i sljedećem broju korisnika:

    Uhhhhh, Cal snagged . Do you have any idea how brilliant that guy is? Lol. And verified lifting isn't just a research toy, it's working for production quality code.

    Poništi
  12. 26. sij

    Thank you to Zoltan Fern, Asif Khan, , and Steven Whang for their help in putting this together, and to everyone who contributed photos or anecdotes.

    Prikaži ovu nit
    Poništi
  13. 26. sij

    A photo tribute to Prof. Hector Garcia-Molina from his Ph.D. students:

    Prikaži ovu nit
    Poništi
  14. 25. sij

    The workshops list is finally up: , w/ 10 workshops spanning HILDA and graphs, to end-to-end ML data mgmt. and db war stories. Thanks to everyone who submitted proposals! Cc:

    Poništi
  15. proslijedio/la je Tweet

    We're sad to share the passing of Prof. Doug Tygar, a long-time professor in the School of Information and . He was a valued community member, teacher, and researcher, and will be sorely missed.

    Poništi
  16. 14. sij

    Are dataframes dying? Maybe not, but many sure are trying to kill them! Wise words from on how we can save dataframes by turning to their historical roots.

    Poništi
  17. 11. sij
    Poništi
  18. proslijedio/la je Tweet
    7. sij
    Odgovor korisnicima

    Evolving tweet. "Stop. A "data frame" is a 4-tuple $(A_{mn},R_m,C_n,Dn)$." If you dataframe, then dataframe with rigor and scale! More in our new paper led by in , and the burgeoning Modin project.

    Poništi
  19. 7. sij

    Led by the awesome , with other coauthors William Ma, , , , Simon Mo, , Anthony Joseph, . Excited to be part of this team! (3/3)

    Prikaži ovu nit
    Poništi
  20. 7. sij

    The Modin open source effort has been ongoing for almost 2 years, and has garnered much industry interest: (see ). Our overall vision is to adapt familiar database optimization techniques to bear on dataframes. Preprint: (2/)

    Prikaži ovu nit
    Poništi
  21. 7. sij

    🎉New preprint! 🎉 Modin is our effort to make dataframes interactive at scale. Modin is a drop-in replacement for pandas, and applies simple optimization techniques to provide up to 30x speedup vs pandas in some cases (map, group-by), and not crash in others (transpose) (1/)

    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:

    ·