Tweetovi

Blokirali ste korisnika/cu @emaadmanzoor

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

  1. Prikvačeni tweet
    11. sij
    Prikaži ovu nit
    Poništi
  2. 23. sij

    Hypothesis: Paper titles induce a buzzword hierarchy that predicts position in a "subtree". My attempt at an ML subtree: /root: "Algorithms for X", "AI for Y" /root/CS: "Self-driving Y" /root/CS/ML: "Deep X", "Hierarchical Y" /root/CS/ML/STATML: "Semiparametrically efficient X"

    Poništi
  3. proslijedio/la je Tweet
    11. stu 2019.

    I’m of the opinion that unless your instrument is ridiculous sounding, and only makes sense once you know the endogenous variable, the exclusion restriction probably doesn’t hold.

    Poništi
  4. proslijedio/la je Tweet
    22. sij

    Our educational paper on visual intuition for influence functions is out in The American Statistician! arXiv PDF: IFs are central to many stat methods and can bridge machine learning with inference, but they're hard to learn (1/2)

    Prikaži ovu nit
    Poništi
  5. proslijedio/la je Tweet
    13. sij

    For the next month or so, I plan to mention an Econometrics paper that I really enjoy reading. I will try to do this every Monday. The first paper of this series is by and , "An Honest Approach to Parallel Trends" Link here:

    Prikaži ovu nit
    Poništi
  6. proslijedio/la je Tweet
    12. sij

    Spoiler alert : "Music is irreplaceable" We used quasi-experimental techniques to show how adopting podcasts influences the original way people listen to music. The paper also includes interesting findings for systems that aim at surfacing multiple mediums simultaneously.

    Poništi
  7. 11. sij

    - Complements the taxonomy curation workflow by + (ISWC 2019: ) with an automated pathway. - Leverages "PinText" multitask embeddings (KDD 2019: )

    Prikaži ovu nit
    Poništi
  8. proslijedio/la je Tweet
    10. sij

    In this new paper we developed a simple, fast and space-efficient representation of a sliding window over an unbounded stream: Age-Partitioned Bloom Filters

    Prikaži ovu nit
    Poništi
  9. proslijedio/la je Tweet
    31. pro 2019.

    manuscript in progress (2̶0̶1̶9̶2020)

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

    After getting published in ICLR as an Independent Researcher, I have received nearly 100 messages from others who are looking to do the same. So I wrote a blog post on why I decided to do it and my advice to others.

    Prikaži ovu nit
    Poništi
  11. 28. pro 2019.

    The library contains a well-documented TMLE implementation in Python (among many other useful tools), accepts arbitrary methods for the ML component: Good complement to the econml library from

    Poništi
  12. proslijedio/la je Tweet
    27. pro 2019.
    Odgovor korisniku/ci

    Thanks for this interesting question. In considering an answer, I realized twitter wouldn’t be the right medium to discuss this in the appropriate level of depth. Instead, here’s a blog post

    Poništi
  13. proslijedio/la je Tweet
    25. pro 2019.

    Try the following exercise: open a registry of randomized trials like the one here: . Try to write down 95% CIs for the effect size before looking at results. You can start to train yourself to see how much uncertainty you should have about the world.

    Prikaži ovu nit
    Poništi
  14. 22. pro 2019.
    Prikaži ovu nit
    Poništi
  15. 22. pro 2019.

    Extensive simulation study of nonparametrics/ML + singly/doubly-robust estimators with a very accessible discussion: by +

    Prikaži ovu nit
    Poništi
  16. proslijedio/la je Tweet
    20. pro 2019.
    Odgovor korisniku/ci

    AC: "I find this style of review entirely inappropriate and unfair: it is not a the role of a good scientific publication to "surprise"." Not all heroes wear capes.

    Poništi
  17. 19. pro 2019.

    Really, really good slides on double machine learning by Chris Felton at Princeton.

    Poništi
  18. 17. pro 2019.

    Super well-documented Python package by for double machine-learning (and many other causal inference methods)

    Poništi
  19. proslijedio/la je Tweet
    16. pro 2019.

    Folks: I made a regression library for Python that can handle high-dimensional fixed effects using sparse matrices. It does interactions, clustering, absorption, the works. With 5m observations and 1k categories it takes about 6 seconds. Check it out at !

    Prikaži ovu nit
    Poništi
  20. proslijedio/la je Tweet
    15. pro 2019.
    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:

    ·