Frank Harrell

@f2harrell

Prof, Founding Chair of Biostatistics, Vanderbilt U. Expert Statistical Advisor, Office of Biostatistics, FDA CDER. Blog: Statistical Thinking -fharrell.com

Nashville, TN
biostat.mc.vanderbilt.edu/FrankHarrell
Dołączył styczeń 2017
154 zdjęcia lub filmy Zdjęcia i filmy

Tweety

Zablokowano @f2harrell

Czy na pewno chcesz zobaczyć te Tweety? Zobaczenie Tweetów nie odblokuje @f2harrell

  1. podał/a dalej
    19 godzin temu

    1. Shut down 2. Jail key decision makers and possibly some of the engineers involved 3. Give the money back to the people it abused (its users) 4. Build ethical replacements for its useful bits

    1 odpowiedź 3 podane dalej 17 polubionych
    Pokaż ten wątek
    Pokaż ten wątek
    Cofnij
  2. 16 godzin temu

    Glad that has restarted this hot discussion on datamethods. I just posted a response that tries to generalize the question.

    W odpowiedzi do to @RogueRad @f2harrell i jeszcze 6osób
    Topic created here:
    3 podane dalej 9 polubionych
    Cofnij
  3. podał/a dalej
    19 gru

    Statistically, machine learning has poor modeling practices which should limit generalizability. It doesn't. Why? For tasks like image recognition, all humans on the ML team understand ground truth; in healthcare/science tasks, it won't be the same. We can expect lots of failure.

    9 odpowiedzi 28 podanych dalej 135 polubionych
    Cofnij
  4. podał/a dalej
    23 godziny temu

    My Christmas present for the community: 2.7 is now on CRAN extending support for Gaussian processes and fixing a substantial performance bug in some post-processing methods among other new features and improvements.

    4 odpowiedzi 71 podanych dalej 233 polubione
    Cofnij
  5. podał/a dalej
    17 gru

    Please read—well worth your time. Challenges of Non–Intention-to-Treat Analyses | Research, Methods, Statistics | JAMA | JAMA Network

    4 podane dalej 17 polubionych
    Cofnij
  6. podał/a dalej
    17 gru

    It's a big stretch to put such a positive spin on the wastage of resources that occurs when the scientific pathway is led astray building upon results that can't be replicated.

    “It’s not a replication crisis. It’s an innovation opportunity” Australian cancer researcher Glenn Begley about reproducibility and the meaning of failure in science
    2 podane dalej 3 polubione
    Cofnij
  7. 17 gru

    Glad to see questionable research practices elevated with respect to severity of harm to science

    Is it time for a new classification system for scientific misconduct?
    3 podane dalej 10 polubionych
    Cofnij
  8. podał/a dalej
    17 gru

    A+ beautiful, a grand achievement by it is wonderful to see Byrne on clean background, all the original copies I saw over the years all had extensive foxing. In my Envisioning Information, we fully rebuilt 2 Byrne pages, hand-separated colors for printing.

    I recreated Byrne's 1847 edition of Euclid's Elements including interactive diagrams, cross references, and a new poster of all the original illustrations: Plus an in-depth blog post on how it was made:
    1 odpowiedź 17 podanych dalej 78 polubionych
    Cofnij
  9. podał/a dalej
    16 gru

    Let me get this straight: Sears has $25 million laying around to pay out executives who ran the company into the ground, but doesn't have any money to give the thousands of workers they laid off severance pay?

    Sears bankruptcy court OKs $25 million in bonuses for top execs
    3 712 odpowiedzi 32 251 podanych dalej 76 942 polubione
    Cofnij
  10. podał/a dalej
    15 gru

    Tonight’s ruling creates confusion, puts affordable health insurance at risk, and targets Americans with pre-existing conditions. The Supreme Court already upheld the law’s constitutionality — twice. Sign up tonight for 2019 coverage at .

    5 odpowiedzi 53 podane dalej 105 polubionych
    Cofnij
  11. podał/a dalej
    14 gru

    "It took more than a half century to reach the point where faced with a choice between democracy and power, the party chose the latter," argues George Packer.

    325 odpowiedzi 5 246 podanych dalej 8 712 polubionych
    Cofnij
  12. podał/a dalej
    14 gru

    Didn’t he impeach Bill Clinton over the same thing?

    Lindsey Graham: I’m fine with Trump "lying about sex"
    7 712 odpowiedzi 34 259 podanych dalej 179 391 polubionych
    Pokaż ten wątek
    Pokaż ten wątek
    Cofnij
  13. podał/a dalej
    14 gru

    A couple days ago another team asked me to speak about Bayesian data analysis. I decided that instead of doing a nuts/bolts of how to fit/use Bayesian models, I would describe "Bayesian analysis: The Good Parts". <potentially controversial thread>

    17 odpowiedzi 231 podanych dalej 814 polubionych
    Pokaż ten wątek
    Pokaż ten wątek
    Cofnij
  14. podał/a dalej
    13 gru

    Not that this is terribly surprising, but still an important message for all to hear.

    0:38
    In the U.S., law enforcement officers kill more than 1,000 civilians each year. New research shows that people who live in states with high gun ownership are much more likely to be shot and killed by police than those in low-gun states
    1 odpowiedź 5 podanych dalej 2 polubione
    Cofnij
  15. podał/a dalej
    12 gru

    Invite others to submit their lists of prevalent faux-pas' compromising reliable evidence generation.

    Value-based Healthcare: analytics may be leaving value on the table
    2 podane dalej 1 polubiony
    Cofnij
  16. podał/a dalej
    12 gru

    Hi everyone. It’s been awhile since I’ve had a good old-fashioned methods critique, but I'm feeling it today:

    4 odpowiedzi 28 podanych dalej 98 polubionych
    Pokaż ten wątek
    Pokaż ten wątek
    Cofnij
  17. 12 gru

    Many Analysts, One Data Set – many conclusions - good discussion continues. Hope that joins in -

    2 odpowiedzi 17 podanych dalej 53 polubione
    Cofnij
  18. 11 gru

    Favorite way to picture probabilities: not with stick figures or frequencies but with a thermometer plot (example below is taken from )

    1 odpowiedź 9 podanych dalej 29 polubionych
    Cofnij
  19. 11 gru

    Statistician take of the day: A $100 data analysis can make a $1,000,000 study worth $1,000.

    4 odpowiedzi 20 podanych dalej 127 polubionych
    Cofnij
  20. 11 gru

    What the GOP stands for ...

    1 odpowiedź 6 polubionych
    Cofnij

@f2harrell nie wysyłał/a jeszcze żadnego Tweeta.

Wydaje się, że ładowanie zajmuje dużo czasu.

Twitter jest przeciążony lub wystąpił chwilowy problem. Spróbuj ponownie lub sprawdź status Twittera, aby uzyskać więcej informacji.

    Może Ci się też spodobać

    ·