DavidSJ

25 ms·

271.
by DavidSJ 6y ago
The issue is not the word "statistical". It's the inversion of P(data | hypothesis) into P(hypothesis | data). I updated my comment to clarify.
272.
by DavidSJ 6y ago
I would guess there are many more than thousands of such experiments that have been run, testing all the various permutations of dozens (or more) of different beyond Standard Model physical theories.
273.
by DavidSJ 6y ago
The measurement from LHCb is three-sigma - meaning there is roughly a one in 1,000 chance that the measurement is a statistical coincidence. No, that’s not what this means. But this is about one in 1,000 BBC Science articles that have rep
274.
by DavidSJ 6y ago
Math from Three to Seven — The Story of a Mathematical Circle for Preschoolers: https://www.msri.org/people/staff/levy/files/MCL/Zvonkin.pdf
275.
by DavidSJ 6y ago
Can you elaborate on the specific issues you see with Road & Track's coverage in this case?
276.
by DavidSJ 6y ago
If you have a steady income and significant savings (say, anything more than six months of expenses), and you haven’t invested a significant fraction of that savings in some growth-oriented asset class like stocks, you’re making a very fool
277.
by DavidSJ 6y ago
What is an example of a string that can be matched which does not meet the first or second description?
278.
by DavidSJ 6y ago
Yes, those are three ways of saying the same thing. :)
279.
by DavidSJ 6y ago
The ∆v to escape Earth orbit is generally quite large, and it's far, far greater still to collide with the Sun (though this wouldn't normally be considered necessary). In comparison, a deorbit burn is usually relatively inexpensiv
280.
by DavidSJ 6y ago
You can measure the generalization gap: the difference between training and test performance. With good generalization that gap will be small. When fitting random labels the gap will be large. Classical statistical learning theory (from Vap
281.
by DavidSJ 6y ago
Random labeling can’t generalize. But in the usual case, the labels are nonrandom and the learned map generalizes.
282.
by DavidSJ 6y ago
The thing is, despite the fact that deep neural networks can effectively learn the hash function you describe (i.e. they have the capacity to do so, and they will do so if "forced" to with random labels), they "prefer"
283.
by DavidSJ 6y ago
We don’t necessarily think that from first principles. But the hypothesis does make certain predictions which are consistent with what we see.
284.
by DavidSJ 6y ago
I think any step which occurred independently on multiple occasions is probably not a hard step. The origin of multicellular life from unicellular life occurred on at least three separate occasions independently. However, I think at the ver
285.
by DavidSJ 6y ago
This is predicted by the hypothesis that the origin of intelligent life requires the occurrence of several hard steps, each of which take at least several billion years (but with no upper bound) in expectation. If that hypothesis holds, the
286.
by DavidSJ 6y ago
I independently had the same thought.
287.
by DavidSJ 6y ago
The existence of human irrationality is not unknown to economists. Behavioral economics, which concerns itself with the effects of this phenomenon on economic decision-making, is one of the major branches of the subject studied today.
288.
by DavidSJ 6y ago
I would have the same question of the US federal government being breached: which systems, exactly?
289.
A 1982 chess computer plays itself by mechanically moving the pieces [video](youtube.com)
3 points by DavidSJ 6y ago|0 comments
290.
by DavidSJ 6y ago
The environmental impact of mining is quite substantial and I agree that this calls into question whether cryptocurrencies are of overall benefit to society. However, it is important to note that mining does do something useful: it secures
291.
by DavidSJ 6y ago
> a particular media company that currently dominates the media landscape here What is this company, out of curiosity? My guess is ABC, but I don't know.
292.
by DavidSJ 6y ago
It might be a good store of value (aside from its extreme volatility), but the GP comment would be correct if you substitute "medium of exchange" for "store of value".
293.
by DavidSJ 6y ago
I think the requirements were calculated at the brokerage level and not the individual customer level.
294.
by DavidSJ 6y ago
That's helpful information. It doesn't invalidate the concern, however, since very large earthquakes are often preceded by moderately large foreshocks, so there is an now a temporarily elevated risk of a very large earthquake in t
295.
by DavidSJ 6y ago
> But where are they supposed to do the more communication? Surely they can't go reading and responding to every thread online that discusses Signal - there's just so many of them. They could put out an official statement on
296.
by DavidSJ 6y ago
Not no communication. Almost no communication. They have occasionally given explanations for why the feature is difficult to implement, and occasionally given explanations for why they think the feature shouldn’t exist at all even if it cou
297.
by DavidSJ 6y ago
FWIW, my experience with Signal sadly confirms this. There’s a critical issue for years with the iOS app that there’s no way to backup or otherwise extract your chat logs (contrary to the usual behavior of iOS apps which automatically backu
298.
UAE Hope Probe Mars Orbit Insertion Live(youtube.com)
1 points by DavidSJ 6y ago|0 comments
299.
by DavidSJ 6y ago
What are all the horizontal bands?
300.
by DavidSJ 6y ago
One specific question about your reference [5]. The article's headline says "Confirmed: Guideline that government will only help ‘51%-black-owned companies’ is fake", although on its face this seems to contract the point I un

More ›