DataScienceWeekly

@DataSciNews

An in-depth look at the world of Data Science. Sign up for our free weekly newsletter featuring curated news, articles and jobs related to Data Science...

datascienceweekly.org
Geregistreerd in november 2013
578 foto's en video's Foto's en video's

Tweets

Je hebt @DataSciNews geblokkeerd

Weet je zeker dat je deze Tweets wilt bekijken? @DataSciNews wordt niet gedeblokkeerd door Tweets te bekijken.

  1. Vastgemaakte Tweet
    12 mrt. 2015

    => Data Scientists at Work >> ... interviews with 16 leaders in the field ...

    1 antwoord 27 retweets 43 vind-ik-leuks
    Ongedaan maken
  2. heeft geretweet
    21 jun.

    Add Constrained Optimization To Your Toolbelt! Our latest post on all things data science. via

    8 retweets 40 vind-ik-leuks
    Ongedaan maken
  3. heeft geretweet
    21 jun.

    Everyone deserves a good data dictionary. Here is my process to build one:

    4 retweets 10 vind-ik-leuks
    Ongedaan maken
  4. heeft geretweet
    21 jun.

    In our most recent collaboration with Google Brain, we measure the similarity between neural network representations to provide insights into generalisation and the training dynamics of RNNs. Blog: Paper:

    1 antwoord 172 retweets 490 vind-ik-leuks
    Ongedaan maken
  5. heeft geretweet
    20 jun.

    Neural ODEs: Instead of updating hiddens layers by layer, we specify their derivative wrt depth with a neural network. An ODE solver adaptively computes the output. By amazing students .

    11 antwoorden 275 retweets 811 vind-ik-leuks
    Deze collectie tonen
    Deze collectie tonen
    Ongedaan maken
  6. heeft geretweet
    23 jun.

    Automatically removing raindrops from an image. Useful for self-driving cars. Trains an attentive generative network using adversarial training. Learns about raindrop regions & their surroundings, then generates focused on surroundings.

    15 antwoorden 459 retweets 1.126 vind-ik-leuks
    Ongedaan maken
  7. heeft geretweet
    19 jun.

    and the whole team at has done a great job with their bite-sized tutorials! See one of our favorites on PyTorch;

    1 antwoord 4 retweets 5 vind-ik-leuks
    Ongedaan maken
  8. heeft geretweet
    22 jun.

    My colleague Raghu has just released "Quantizing deep convolutional networks for efficient inference" - This white paper covers practical quantization approaches for most common CNNs in , I think it will be super useful!

    1 antwoord 69 retweets 172 vind-ik-leuks
    Ongedaan maken
  9. heeft geretweet
    19 jun.

    Featuring also

    How AI and robotics are changing health care - our discussion:
    1 retweet 5 vind-ik-leuks
    Ongedaan maken
  10. heeft geretweet
    20 jun.

    Writing image processing algorithms to count cell nuclei was one of my summer jobs as an undergraduate. Great to see deep learning applied to this problem:

    Take Two Algorithms and Call Me in the Morning. Bowl winners use on for spotting cell nuclei in dense images.
    1 retweet 2 vind-ik-leuks
    Ongedaan maken
  11. heeft geretweet
    20 jun.

    Nature just published a major feature on researchers working on bias in machine learning. Features many of us, including - and the work on due process

    11 antwoorden 499 retweets 796 vind-ik-leuks
    Ongedaan maken
  12. heeft geretweet
    19 jun.

    Building a real world, production machine learning application with and - search guru dives into how we built

    9 retweets 12 vind-ik-leuks
    Ongedaan maken
  13. heeft geretweet
    19 jun.

    Best paper award at main idea: study twenty five different visual tasks to understand how & when transfer learning works from one task to another, reducing demand for labelled data. Paper: Data:

    2 antwoorden 480 retweets 1.267 vind-ik-leuks
    Ongedaan maken
  14. heeft geretweet
    19 jun.

    Cool, beautiful, and informative. What more could you ask for?

    0:40
    In any machine learned model, including neural networks, one of (1) erroneous assumptions in the learning algorithm, or (2) sensitivity to small fluctuations in the training set, is unavoidable. A beautifully crafted explanation of bias-variance tradeoff:
    1 retweet 3 vind-ik-leuks
    Ongedaan maken
  15. heeft geretweet
    19 jun.

    This AI program could beat you in an argument—but it doesn’t know what it’s saying

    7 antwoorden 91 retweets 137 vind-ik-leuks
    Ongedaan maken
  16. heeft geretweet
    20 jun.

    "Deep Learning for Dialogue Systems" - great [TUTORIAL] by

    12 retweets 55 vind-ik-leuks
    Ongedaan maken
  17. heeft geretweet
    18 jun.

    Happy to share my last project🤗 Metacar: A reinforcement learning environment for self-driving cars 🚗 in the browser. - Solve problems 🔨 - Edit your levels ✍ The github: The demo includes solved problems made with js

    11 antwoorden 126 retweets 371 vind-ik-leuks
    Deze collectie tonen
    Deze collectie tonen
    Ongedaan maken
  18. heeft geretweet
    18 jun.

    A look inside Uber Eats' search and recommendation engine to make sure you aren't frustrated when you get the munchies | by

    4 retweets 5 vind-ik-leuks
    Ongedaan maken
  19. heeft geretweet
    18 jun.

    New approximate nearest neighbor benchmarks (blog post)

    18 retweets 69 vind-ik-leuks
    Ongedaan maken
  20. heeft geretweet
    15 jun.

    I'm fascinated by archaeology, so I'm excited to see trying out the for Poets tutorial on pottery shards!

    1 antwoord 7 retweets 24 vind-ik-leuks
    Ongedaan maken
  21. heeft geretweet
    12 jun.

    ConvNet outperforms human dermatologists for melanoma detection. Dermatologists in level-II protocol: - sensitivity: 88.9% (±9.6%, P = 0.19) - specificity: 75.7% (±11.7%, P < 0.05)....

    9 antwoorden 185 retweets 459 vind-ik-leuks
    Ongedaan maken

@DataSciNews heeft nog niet getweet.

Het laden lijkt wat langer te duren.

Twitter is mogelijk overbelast of ondervindt een tijdelijke onderbreking. Probeer het opnieuw of bekijk de Twitter-status voor meer informatie.

    Je bent misschien ook geïnteresseerd in

    ·