dguest

46 ms·

511.
by dguest 7y ago
I was going to suggest Yr as well! I live in France and it seems to work reasonably well.
512.
by dguest 7y ago
You also need to consider that when the false positives get over their flu or cold or whatever, they will assume they are immune to COVID.
513.
by dguest 7y ago
You should account for ignorance in public health policy, but you shouldn't rely on it. Having more false positives than true positives only works if the people taking the test either don't understand statistics or don't kn
514.
by dguest 7y ago
If you rush a vaccine that causes 5% of patients develop a fatal complication in 1 year you've just killed twice as many people as COVID-19 would have. Personally I'd rather take my chances with COVID-19 than risk injecting myself
515.
by dguest 7y ago
It's interesting that you lead by saying "as a physicist". As a physicist I'm used to thinking about orders of magnitude and whether something grows like log(N) or N!. I'm pretty sure all the data points to this bei
516.
by dguest 7y ago
Maybe to put a more positive spin on it: the sooner you act the more leeway you have. Better that you realize now that someone has to come in and install a remote power switch on some server. In a week it might be a bit harder to ask someon
517.
by dguest 7y ago
> all milk in France is shelf-stable until opened As someone currently in France I thought this was funny: the first thing I show visitors is the "distributeur de lait frais" down the street, which will dispense 1L of raw milk
518.
by dguest 7y ago
We might not use that any more, but unfortunately the concept of lighting one's enemy on fire hasn't gone away. What about flamethrowers? Napalm? Is there any indication that Greek fire is somehow more destructive than the modern
519.
by dguest 7y ago
On one of the bigger experiments we use quite a few things: - AFS as a federated posix file system for user home directories. My impression is that a distributed posix filesystem is... well... hard, for basically the reasons listed in the l
520.
by dguest 7y ago
When you work for an international organization DST is doubly annoying. Four times a year we have 1/4th of our team missing half the meetings: - When Europe goes to DST, half the Americans don't show up to meetings on European tim
521.
by dguest 7y ago
Having work start at 5 am doesn't make time unofficial, does it? It would seem like it would make starting work at 8 am non-standard, but I don't see any logistical or biological reason that people have to go to work when a clock
522.
by dguest 7y ago
Well when you do your PhD in particle physics it's just sort of accepted. And I think it's tough on a lot of people: many people in the experiment feel that they should read every paper, because they feel personally accountable to
523.
by dguest 7y ago
At the LHC the number is closer to 3500. I "cruise" on around a hundred papers each year: I haven't read them and in all likelihood never will. At the same time I worked really hard on one or two of them. I have tried to figu
524.
by dguest 7y ago
To be fair, she's not telling Zuckerburg how to run Facebook. She's saying an external entity (with a very different mandate) should regulate Facebook.
525.
by dguest 7y ago
On the other hand: I'm a bit surprised at how many people are writing JSON and care about the performance. The overwhelming majority of our IO is reading / writing data files, but those are stored in an optimized binary format. Th
526.
by dguest 7y ago
There's a pretty sharp migration away from ROOT to numpy / pandas these days (at least in the younger generation).
527.
by dguest 8y ago
Don't get your hopes up. I worked in a neurology lab where they used something similar years ago. You could keep a pig / rat brain running for a few hours like this, but it would start to fall apart in other ways pretty fast.
528.
by dguest 8y ago
There are limits to how much you can give researchers a pass though. There are plenty of senior researchers who have perfectly secure careers but want a bigger grant, more postdocs, a more prestigious position, etc etc, and of course they h
529.
by dguest 8y ago
I often wish large experiments would take a stronger stance on the supply end. At CERN we've worked out deals with publishers mandating that _our_ content be open access, even if the journal isn't. Personally I feel like this is b
530.
by dguest 8y ago
What's so sad about someone who doesn't really like to spend time outside? Personally I like going outside, but I'm a bit jealous of people who love their work environment that much.
531.
by dguest 8y ago
This is a misleading title. The researcher they quote is > ... developing the next generation of machine learning and statistical techniques that can ... also report how uncertain their results are and their likely reproducibility. So sh
532.
by dguest 8y ago
90% of the work in LHC physics is estimating the amplitude of backgrounds that look almost exactly like your signal process. Coming up with the "large booster" for classification is only a small part of it. So yes, machine learnin
533.
by dguest 8y ago
This is a very good point. I work in an organization that just switched to monolithic and it's been going very well, with hundreds of active developers and millions of lines of code. But our developers are students or academics. As man
534.
by dguest 8y ago
As others have commented we've been using tools like neural networks and boosted decisions trees for a long time. We have quite good simulation which tells us what particles would look like in our detector, but one thing our simulation
535.
by dguest 8y ago
She's not talking about TMVA or RooStats when she says "Machine Learning": those would be "MVA" and "Statistics tools" in our jargon. She's talking about XGBoost[1]. [1]: https://xgboost.re
536.
by dguest 8y ago
I was a little lost at the first paragraph (we do have a name for the strategy behind the "no one ever got fired for buying an IBM" pitch[1]). The next point is just weird for someone who knows anything about academic funding: may
537.
by dguest 8y ago
One of the articles cited there is an excellent article that Edward Epstein wrote for The Atlantic [0]. It's well written, but that's not the impressive part. The impressive thing is that this "expose" was written in 198
538.
by dguest 9y ago
This is a fun read on wikipeida [1]. I'm curious as to how you could disprove it at all: there's a perfectly valid coordinate transformation that maps our "infinity" to the center of the earth, and the center of the eart
539.
by dguest 9y ago
Personally I think a billion dollars for a nationally recognized electronic payment system is worth it. In the US the commercial solutions (venmo, paypal, etc) work fine between friends, but sometimes you get stuck having to write (paper) c
540.
by dguest 9y ago
I think you'd either want the full month name or all the months to be abbreviated to the same number of characters. You can't do 2 characters because of June and July, but three works. 4 would be sort of weird.

More ›