Tweetovi

Blokirali ste korisnika/cu @mandyfmejia

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

  1. Prikvačeni tweet
    31. sij

    Graduating with skills in and in 2020? I'm recruiting a to develop/validate/apply cutting-edge Bayesian statistical methods for . Tools that make it possible to precisely map individual brains w/out huge amounts of data.

    Prikaži ovu nit
    Poništi
  2. 31. sij

    Learn more at or email me at afmejia@iu.edu.

    Prikaži ovu nit
    Poništi
  3. 31. sij

    Excited for the first lab meeting of the semester with new IU Informatics faculty member Wei Tang!

    Poništi
  4. 14. sij
    Poništi
  5. What's your favorite open access dataset with fMRI? So far I have HCP, Baby Connectome, My Connectome, NKI Rockland sample, ABCD study, PNC, UK biobank, 1000 Functional Connectomes, Midnight Scan Club, ADNI and ABIDE. Extra points if it focuses on a specific population!

    Poništi
  6. proslijedio/la je Tweet
    21. stu 2019.

    Bayesian Template ICA approach leads to dramatic (75%-250%) improvements in intra-subject reliability compared to Dual Regression in both simulation and HCP!! 🤩🧠🧠🤩 Very exciting work - looking forward to digging into the code base!

    Poništi
  7. Our paper on reliable subject-level brain network estimation using template ICA is officially out! First 50 eprints are free Matlab toolbox and R package at

    Poništi
  8. proslijedio/la je Tweet
    6. stu 2019.

    Hey there (bio)stat grad students: PLEASE make yourself an academic website by the time your 1st paper is out (if not sooner)! 1/4

    Prikaži ovu nit
    Poništi
  9. The happy end result of a long but rewarding journey with & . Template ICA estimates RSNs in individual subjects reliably using ICA with empirical group priors. And we have a toolbox! . R package in the works too!

    Poništi
  10. proslijedio/la je Tweet
    15. lis 2019.

    I'm recruiting PhD students to join my new lab ! Focus on individual differences under naturalistic conditions using neuroimaging, behavior, and computation. Happy to chat at ! (Lab manager position available too, stay tuned for formal announcement)

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

    Don't forget: will be speaking at Luddy at 1:30 p.m. tomorrow! @IUSICE Can't make it? IUNI will post a livestream link before the event here:

    Poništi
  12. proslijedio/la je Tweet
    3. lis 2019.

    On why the same authors published reanalyses of their prior studies with different interpretations: "To get published, you can't say the same thing. You need to say that you are advancing science, and advancing science means advancing confusion." - John Ioannidis

    Poništi
  13. Also super impressed that Science Correspondent is sticking around for the talks today after his keynote last night. Academics take note! We should put as much attention into listening as we do speaking. We are not too busy.

    Prikaži ovu nit
    Poništi
  14. proslijedio/la je Tweet
    3. lis 2019.

    Forensic bioinformatics talk by John Baggerly... How the heck did they get from data to conclusion? "I'm a data parasite!". Ovarian cancer data and problems with experimental design.

    Prikaži ovu nit
    Poništi
  15. A highly entertaining yet sobering reminder about the Potti Duke cancer trial scandal and the importance of reproducibility in the study of health. If you've never heard about it, the 60 minutes episode is a good place to start.

    Prikaži ovu nit
    Poništi
  16. Favorite quote of the day so far: "Can we infer what the hell they must have done to get from here to here?" -- Keith Baggerly on inventing forensic bioinformatics.

    Prikaži ovu nit
    Poništi
  17. Spending the day at the Rigor Reproducibility & Transparency workshop organized by David Allison at . Very cool mix of thinkers including , Keith Baggerly & John Ioannidis. Love this small and interdisciplinary format!

    Prikaži ovu nit
    Poništi
  18. proslijedio/la je Tweet
    23. ruj 2019.

    We a to work on + funded projects on +. DM/email me if you have questions. Collaborators: Share widely!

    Poništi
  19. We are making the ID-linked OP data publicly available.

    Prikaži ovu nit
    Poništi
  20. I forgot to mention: this was only made possible by manually looking up the national ID of each provider in the Open Payments dataset (by Jorge Mejia). The OP system excludes it, so it's tough to link to other data sources like Medicaid/Medicare.

    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:

    ·