Arinbjörn

@arinbjornkol

PhD student working on ML, tensors, brain MRI and other medical data. Former research intern AI Cambridge

London
Vrijeme pridruživanja: svibanj 2010.

Tweetovi

Blokirali ste korisnika/cu @arinbjornkol

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

  1. Prikvačeni tweet
    5. ožu 2019.

    My internship at Samsung AI: Stochastically Rank-Regularized Tensor Regression Networks Improved performance. Reduced overfitting. More robust to adversarial attacks and random noise. Work with , Yannis Panagakis, et al.

    Poništi
  2. proslijedio/la je Tweet

    how to mess with your coworkers

    Poništi
  3. 2. velj

    The technology equivalent of selling scissors in sealed frustration packaging

    Poništi
  4. proslijedio/la je Tweet
    31. sij

    A full build of Autopilot neural networks involves 48 networks that take 70,000 GPU hours to train. Together, they output 1,000 distinct tensors (predictions) at each timestep. This is what a Tesla autopilot sees [source: ]

    Poništi
  5. 31. sij

    Guidance from on the use of masks Takeaway: Wear a mask as soon as any symptoms appear. Not required for non-infected individuals. Instead stay at least 1 m from anyone with symptoms + wash your hands

    Poništi
  6. proslijedio/la je Tweet
    28. sij

    Journal: Ready to submit your manuscript? Me: Yup! (cracking my knuckles) Journal: Do you have all these ready? 1. Your login details 2. Your ORCID ID 3. Manuscript 4. Each image as separate file 5. Cover letter 6. COI for each author 7. Author signatures for all

    Prikaži ovu nit
    Poništi
  7. 27. sij

    Record levels of inflation, likely due to magma accumulation, around Mt. Þorbjörn If this turns into a eruption (unlikely) it might be bye bye Blue Lagoon

    Poništi
  8. 24. sij

    Things I can do in 6 days: read a book Things China can do in 6 days: build a 1000-bed hospital

    Poništi
  9. 21. sij
    Poništi
  10. 15. sij

    Sequential manufacturing is built to blueprints for outdated human factories that required worker repetition, according to who are betting on parallelised multi-cell factories where each cell produces a vehicle from scratch – mjesto: Kensington Village

    Poništi
  11. proslijedio/la je Tweet

    We are deeply saddened at the tragic news that our PhD student Saeed Tahmasebi Khademasadi was among the passengers to have died in the plane crash in Iran this morning. 1/3

    Prikaži ovu nit
    Poništi
  12. 6. sij

    Reference: Smith, Aaron, and Oscar Sugar. "Development of above normal language and intelligence 21 years after left hemispherectomy." Neurology 25.9 (1975): 813-813.

    Prikaži ovu nit
    Poništi
  13. 6. sij

    The plasticity of the human brain is remarkable. Consider hemispherectomy, a rare operation where half the brain is removed Despite the sacrifice there are cases of patients developing "above normal language and intelligence" Some would say the brain is overparameterized

    Prikaži ovu nit
    Poništi
  14. 31. pro 2019.
    Poništi
  15. proslijedio/la je Tweet
    Poništi
  16. 21. pro 2019.

    Reusable electronic bag tag by auto-updates to match your next flight

    Poništi
  17. proslijedio/la je Tweet
    20. pro 2019.

    UMAP is often said to be superior to t-SNE because it is better at preserving global distances. Dmitry and I showed that this is because by default, t-SNE initializes randomly whereas UMAP initializes with Laplacian eigenmaps.

    Prikaži ovu nit
    Poništi
  18. proslijedio/la je Tweet
    17. pro 2019.

    Want to catch up on some latest Deep RL work? Here is a zip file with all the camera-ready papers from the Deep RL Workshop!

    Poništi
  19. 13. pro 2019.

    I'll be presenting a poster here tomorrow with on a new sample-efficient architecture, where deep RL meets tensor methods and wavelet scattering!

    Poništi
  20. 13. pro 2019.

    I'll be giving a talk with at the Graph Representation Learning workshop at 10:00 in West Exhibition Hall A. Stop by to learn more about the challenges of leveraging time in graphs

    Poništi
  21. 12. pro 2019.
    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:

    ·