pyduan

17 ms·

91.
by pyduan 13y ago
I think there is a fundamental difference you're missing between the experiences you're describing and that of online dating: it's not so much about the amount of noise and more about the signal-to-noise ratio, as well as
92.
by pyduan 13y ago
> This is bollocks, for the most part. There have been plenty of studies... my state was doing a lot of research... studies here and there would filter through... this is not reliant on a single study with a small n. There is a missing p
93.
by pyduan 13y ago
This is a comment that often comes up when the sample size doesn't seem very big, but keep in mind that the required sample sizes to draw statistically significant conclusions can be surprisingly small especially when the effect size i
94.
by pyduan 13y ago
It looks like what they've done is more along the lines of finding a clever trick to improve the runtime performance of one of the common building blocks of a spell correction system (the part that given an input computes a set of ca
95.
by pyduan 13y ago
Remember that both Easterlin's studies and the one in this article have been done in many countries and not just the US though. The $128k cap may seem limiting for the US but it is not for Nigeria or the other lower-income countries,
96.
by pyduan 13y ago
This is precisely what Easterlin says: a relative increase in income has little effect on happiness past a certain point. This competing model (the one in the article) says that no, there is no saturation point and there is a logarithmic re
97.
by pyduan 13y ago
No, the models proposed by Easterlin is a fundamentally different model of happiness economics than the one proposed by the researchers in the article (I updated my comment to flesh this out a bit). Basically, Easterlin's research hint
98.
by pyduan 13y ago
This is one of the big controversies in happiness economics. I wonder if this effect (which directly contradicts Easterlin's findings that money has no effect on happiness past a certain point; here, they contend that there is no sat
99.
by pyduan 13y ago
You're missing my point though. My point was not that opening links in a new window is always a good thing, just that usability experts keep parading the same old studies and scientific-sounding claims about "cognitive load"
100.
by pyduan 13y ago
In this case: the federal government. The banks didn't randomly decide they were against marijuana, they're just reluctant because while marijuana is legal for medicinal or recreational use in a few states it's still a Schedu
101.
by pyduan 13y ago
> Not sure what the point would be… "In God we trust, all others must bring data." Best practices should be grounded in data rather than opinions whenever possible, especially when they're based on claims about users'
102.
by pyduan 13y ago
Are there any more recent studies with actual numbers and not just some vague talk about increased cognitive load (whatever this means)? Most I found seem to refer to the same 15 year old NN/g research report. Since then a lot has chan
103.
by pyduan 13y ago
Thank you for the precision about Apple products. I may be wrong about the two finger gesture being disabled by default (edited my post to reflect that) -- it was my impression that it wasn't, but regardless I think we can agree it is
104.
by pyduan 13y ago
I used to agree with the author, but these days I find myself actually preferring having a link opening in a new tab, so I'm not actually that irritated when a page uses target="_blank" if it's somewhat justified (ie. if
105.
Intel renames its McAfee security brand(money.cnn.com)
3 points by pyduan 13y ago|0 comments
106.
Random Forest(en.wikipedia.org)
4 points by pyduan 13y ago|0 comments
107.
You are your brain - Why is neuroexistentialism so disturbing?(slate.com)
2 points by pyduan 13y ago|0 comments
108.
by pyduan 13y ago
This is called ad retargeting -- the basic premise being that conversion rates are much higher for items the user has actually looked at in the past. Usually it's only supposed to show abandoned purchases, but sometimes it doesn't
109.
by pyduan 13y ago
I initially had the same reaction as you, but reading the discussion between the author and Telegram in the comments (both on the linked article and the original HN thread) it seems to be this is largely deserved given the way Telegram has
110.
Emotional intelligence predicts job success: do you have it?(fastcompany.com)
1 points by pyduan 13y ago|0 comments
111.
by pyduan 13y ago
It's quite sad that this post is even necessary. That said, having a proper training/cross-validation/validation setup is sometimes not that obvious, as you have to stop and think about possible sources of contamination -- so
112.
by pyduan 13y ago
True, but intuition and general analytical ability is often overlooked when hiring data scientists -- because the term is so broad, what people are often looking for is a mixture of statistical know-how and programming ability, but often fo
113.
The data is not enough: Creative data scientists make the difference(venturebeat.com)
15 points by pyduan 13y ago|2 comments
114.
by pyduan 13y ago
On a semi-humoristic note, there's also John Perry's Structured Procrastination theory, which won him the 2013 Ig Nobel prize in literature: http://www.structuredprocrastination.com/ Basically, be productive by do
115.
by pyduan 13y ago
The fact you use Jesse Owen's example is a bit ironic, since he was actually much better treated in Germany than in the United States where racial segregation was still the norm. By his own account: "Hitler didn't snub me – i
116.
by pyduan 13y ago
The fact there are arbitrage opportunities are precisely the reason the regulations exist. The problem is that the market equilibrium would be really bad for residents. You have to stop and think about what there being no such opportunity
117.
by pyduan 13y ago
I know Indiegogo is hard at work towards building automated scam/fraud detection systems. I work in this space (I build such systems for Eventbrite) and from my conversation with them it seems they're new to these problems but are
118.
by pyduan 13y ago
Citation needed. 1/ There are gullible (not necessarily dumb, by the way) people everywhere, why would Kickstarter's user base be exempt from that? Besides, Kickstarter is fairly mainstream now, so there's no reason to think
119.
by pyduan 13y ago
Kickstarter is not the one who needs the convincing at this stage, it's the potential backers. Case in point, this post does not link to an official Kickstarter investigation, but to a spontaneous one made by a random commenter.
120.
by pyduan 13y ago
That reminds me of that argument that scammers purposefully introduce spelling mistakes and other types of obvious red flags because it filters out the smart recipients and only leaves people who are likely to be gullible enough to take the

More ›