Rezultati pretraživanja
  1. 8. pro 2018.

    All of data science in one slide :p

    Prikaži ovu nit
  2. 13. pro 2018.
  3. 16. pro 2018.

    GAN Demystified, WTH do they learn. An article from @henddkn who I had the chance to meet during

  4. Kind of unbelievable that Yo-yo Ma is performing at

    Prikaži ovu nit
  5. 7. pro 2018.

    It's amazing to see how is helping people to attend remotely. She is a great example to others. Thanks to and for organizing the workshop.

  6. 20. pro 2018.

    My slides for the Meta-Learning are now up. Big thanks to the organisers!

  7. 11. pro 2018.

    Code for our paper "RenderNet: A deep convolutional network for differentiable rendering from 3D shapes" can be found here: . Trained models and data to follow :)

  8. After seeing 33% of African citizens denied visas to attend , and knowing that 15% hear back too late to attend, considerations like these from conferences are *huge*. Talent is global; the open-source community is global.

    Prikaži ovu nit
  9. 17. pro 2018.

    All the slides for the talks at the MLOSS workshop are now online:

  10. 8. pro 2018.

    Slides for my talk on "Probabilistic model ensembles for predictive uncertainty estimation" at the Bayesian deep learning workshop @ are available at Check out the deep ensembles and Nature Medicine papers (links in slides) for more details :)

  11. 7. pro 2018.

    Explaining our work on incremental Boltzmann Machines to Geoffrey Hinton, the inventor of Boltzmann Machines.

  12. 8. pro 2018.

    Important reminder from talk at : respect other fields (see also: don’t oversell)

  13. 7. pro 2018.

    Geoffrey Hinton is such a humble and humorous person! Finally, got the chance to hear his thoughts about our work on marrying Boltzmann machines and deep generative learning

  14. 12. pro 2018.

    The AI art gallery from our NeurIPS Creativity workshop is now online 😀🤖

  15. 10. pro 2018.

    My summary from music x ML (=music information retrieval (=MIR)) perspective is here:

  16. 11. pro 2018.

    If you missed our paper at on training neural nets straight from JPEG representations, this blog gives a quick summary. Or even shorter: "Image classification is faster AND more accurate from JPEG!"

  17. Based on the feedback of newsletter readers & another sit-down with the ever-generous , I have a new & improved explainer on the best paper of . I think my brain muscles doubled in size from the workout of getting this right. 💪🧠

    Prikaži ovu nit
  18. 19. pro 2018.

    The Borealis AI Skynet team recently took 2nd place @ the team competition. Learn how they set up for battle and rose to the top in this challenging multi-agent environment. (Bonus if you can name the film reference in the title):

  19. 11. pro 2018.
    Prikaži ovu nit
  20. 17. pro 2018.

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.