David Josephs

@Daved256

Pea-brained data scientist looking for a job in may! Interests: Explainable AI, Unsupervised Learning, forecasting, python, R, and Nix Github: josephsdavid

Vrijeme pridruživanja: listopad 2019.

Tweetovi

Blokirali ste korisnika/cu @Daved256

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @Daved256

  1. proslijedio/la je Tweet
    27. sij

    The fun part about this article is that it is possible to read this in any of three ways: as a lesson on false positives/false negatives, a guide for how to improve your process for selecting candidates, or a long-form diatribe about job hunting.

    Prikaži ovu nit
    Poništi
  2. proslijedio/la je Tweet

    What Michael was to Kobe, Kobe became to the next generation of players.

    Poništi
  3. proslijedio/la je Tweet
    26. sij

    My first blog post attempt. 🥳 I tried to intuitively explain t-SNE and show the effects of different hyperparameters on the visualization.

    Prikaži ovu nit
    Poništi
  4. proslijedio/la je Tweet
    25. sij

    gpt-2 is better at writing cover letters than I am

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

    Honestly I'm tired of coding in Python. It's very snobby and elitist. I'm going back to my roots and doing everything in Microsoft Excel from now until further notice. Excel is data science software for the common man.

    Tweet je nedostupan.
    Poništi
  6. proslijedio/la je Tweet
    21. sij

    Statistics: hey Machine Learning, what did you get on the physics test? ML: 100% Statistics: lol. I got 80 ML: lol. I just crammed all the practice tests Statistics: can you help me with this homework? ML: sure. um. quick question though Statistics: yeah? ML: what’s gravity?

    Poništi
  7. proslijedio/la je Tweet
    20. sij

    Given the smoothness of videos, can we learn models more efficiently than with ? We present Sideways - a step towards a high-throughput, approximate backprop that considers the one-way direction of time and pipelines forward and backward passes.

    Poništi
  8. 11. sij

    Lots of really interesting stuff here

    Prikaži ovu nit
    Poništi
  9. 11. sij

    This is a thread worth reading and saving

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

    A tour de force by Abbe & Sandon, "Any function distribution that can be learned from samples in poly-time can also be learned by a poly-size neural net trained with SGD on a poly-time initialization with poly-steps" + "[this] does not hold for GD"

    Poništi
  11. proslijedio/la je Tweet
    9. sij

    On Interpretability of Artificial Neural Networks by Fenglei Fan et al. including

    Poništi
  12. proslijedio/la je Tweet
    9. sij
    Poništi
  13. proslijedio/la je Tweet
    7. sij

    this is absolutely nuts. The AI GPT-2 has learned to play chess moderately well (able to give bad human amateurs a game) – despite only being a text AI, learning from a corpus of chess notation text, and not having any concept of what a chessboard is

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

    top places i *do not* want to be smart: - my body - toilet - bathroom

    Poništi
  15. proslijedio/la je Tweet
    5. sij

    Consider: millions of years ago our antecedents gave a massive sacrifice of their left hemisphere. We lost a tremendous amount of short term memory and replaced it with Broca’s, Wernicke & the phonological loop. But why? So we can—talk. Thus chimpanzees can do this—we can’t:

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

    Pynndescent, an approximate nearest neighbor search library, got a major update recently. Index construction is now multicore by default. Querying is now much faster -- competitive with some of the fastest ANN libraries around. (1/4)

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

    What are some of the best-written papers you've come across? My favorite so far: Computing Machinery and Intelligence by Alan Turing. A great example of complex ideas made easy to grasp: Turing explained intelligence using party games!

    Poništi
  18. proslijedio/la je Tweet
    4. sij
    Prikaži ovu nit
    Poništi
  19. proslijedio/la je Tweet

    Random thought: Articles on should stop focusing on futuristic self-driving trolley cars as examples (which still need years) and instead put more focus on existing concerning real-world use cases such as COMPAS and PredPol. It would help raise awareness of awful AI❗

    Poništi
  20. proslijedio/la je Tweet
    2. sij

    Let’s talk about what’s happening in Australia, which is suffering its worst bushfire season in recorded history. This almost unimaginable catastrophe has killed half a billion animals and released over 250 million tonnes of CO2 so far.

    Ovo je potencijalno osjetljiv multimedijski sadržaj. Saznajte više
    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:

    ·