Savage Jim

@jim_savage_

Research technologies . Thoughts on stats, econ and philanthropy from this account. Sillier stuff at

New York, USA
Vrijeme pridruživanja: ožujak 2012.

Tweetovi

Blokirali ste korisnika/cu @jim_savage_

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

  1. Prikvačeni tweet
    26. tra 2018.

    - Highly significant effect - Fixed cost of implementing treatment - Diminishing returns on loss function to treatment - Evaluating your loss function at expected TE tells you the wrong thing to do. ABIYLFOYP Always Be Integrating Your Loss Function Over Your Posterior

    Poništi
  2. prije 54 minute

    Imagine that we privatized non-primary streets. Local + service vehicles only, bollarded. Residents would love it, but it would make driving impossible. I wonder if there’s a lesson here about the role of private groups in the death of Facebook? Will DM groups kill Twitter?

    Poništi
  3. prije 2 sata

    Great plots too. Are there any non-academic think-tanks anywhere that produce work like this? Demos or Brookings, maybe?

    Poništi
  4. proslijedio/la je Tweet

    Super bowl sponsorship by Google - are people like “yeah, maybe I will google something”

    Poništi
  5. 2. velj

    Saw Uncut Gems tonight. Fantastic film. Great story, but also technically brilliant. Casting, costumes, lighting, sound, music: all 5/5. Just delicious.

    Poništi
  6. proslijedio/la je Tweet
    1. velj

    Car-free zones I'd like to see in NYC, LA, SF, and Boston. Could start with "only local residents and workers get a car permit to enter the area", and speed limit is very low, to begin with.

    Poništi
  7. 1. velj

    In my world (expat Australians in tech/econ/policy), because of his willingness to have very long lunches with interesting strangers followed by dozens of follow-up emails sharing essays.

    Prikaži ovu nit
    Poništi
  8. 1. velj

    Does your community have any people that everyone seems to consider a mentor or coach? Who are they, and what makes their approach special?

    Prikaži ovu nit
    Poništi
  9. 1. velj

    Useful ! 1- nice blog post. 2- sometimes you’re on a public WiFi and it won’t bring up the login window. In those cases, you can usually prompt the login window to pop up by navigating to a non-https website, but there aren’t many left.

    Poništi
  10. 1. velj

    On reporting rates--this is a real challenge! I'm sure there are clever people who do this for a living, but I'd use plug-in rates for sensitivity checking. It just doesn't seem like a easy thing to estimate. Have a great weekend!

    Prikaži ovu nit
    Poništi
  11. 1. velj

    On dist choice: one nice fact is that a mixture of normal densities has a mixture of normal CDF. So you can use mixtures of normals for your f()s to capture most of the weirdness going on in unusual arrival time distributions, while staying analytically convenient.

    Prikaži ovu nit
    Poništi
  12. 1. velj

    log_sum_exp(a, b) = log(exp(a) + exp(b)). So log(θ*(1 - F(d|t))+ (1 -θ)*(1-F(c|t))) = log_sum_exp(log(θ) + log(1 - F(d|t)), log(1-θ) + log(1 - F(c|t))). Beautiful! Two more challenges: what distributions for f(d|t) and f(c|t); and what about low reporting?

    Prikaži ovu nit
    Poništi
  13. 1. velj

    If they're neither dead nor alive, a bit more tricky. Then it's the probability they're going to die but haven't yet, plus the probability they'll live but haven't been cleared yet. So θ*(1 - F(d|t))+ (1 -θ)*(1-F(c|t)). Taking logs of that is a bit nasty, so we use log_sum_exp

    Prikaži ovu nit
    Poništi
  14. 1. velj

    So what's the likelihood of observing this data for a patient, p(s,t|θ)? If they're dead, it's θ*f(d|t), so log likelihood contribution is log(θ) + log(f(d|t)) If they're cleared, it's (1-θ)*f(c|t); ll contribution = log(1-θ) + log(f(c|t)).

    Prikaži ovu nit
    Poništi
  15. 1. velj

    Let's call f(d|t) the density of death arrival times t. f(c|t) is the density of dates that you get cleared. F() indicates CDF. θ is the death rate to be estimated. The data we need is each patient's state s (dead, sick, cleared), and days since infection, t.

    Prikaži ovu nit
    Poništi
  16. 1. velj

    You wouldn't say that it's 100/200= .5. There are all these sick people who haven't died or ben cleared yet. Their survival so far contains useful information for our estimate. So how do we use it?

    Prikaži ovu nit
    Poništi
  17. 1. velj

    For the moment, let's set reporting rates to 1. Why is the arrival time of death vs recovery an issue? Imagine the death rate is 20%. Deaths take 1 week on average, but rec takes 3 weeks. You infect 1000 people and wait two weeks; 100 die, and 100 are cured. What's your estimate?

    Prikaži ovu nit
    Poništi
  18. 1. velj

    By the time the illness has passed, we could estimate its death rate given infection as (reported deaths + unrep'd deaths)/(rep'd deaths + unrep'd deaths + rep'd recoveries + unrep'd recs). 2 problems for us: unreported X isn't known, and deaths might arrive sooner than recovery

    Prikaži ovu nit
    Poništi
  19. 1. velj

    It's a Friday night, so a good time to derive the likelihood function of a simple model to estimate the death rate from an illness that kills quickly but takes a long time to cure. Buckle up!

    Prikaži ovu nit
    Poništi
  20. 31. sij

    Am I reading this right? Are 6-8% of US deaths due to flu and pneumonia?

    Poništi
  21. 30. sij

    An Australian social institution I’d like to introduce: the 6:30/7am breakfast catch-up. You meet someone at a conference and they’re passing through town? Totally acceptable to propose a 7am breakfast at a nice cafe. No alcohol, great food, and a natural end in case it’s awk.

    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:

    ·