lars

10 ms·

61.
On the Influence of the Sewing-Machine on Female Health (1867) [pdf](ncbi.nlm.nih.gov)
1 points by lars 11y ago|1 comments
62.
by lars 11y ago
> the high terrorist attack output of Islamists vs other non-Islam subscribers I wonder if this is really true though. Some statistics: https://www.fbi.gov/stats-services/publications/terrorism-20... While Ande
63.
Focusing Illusion (2011)(edge.org)
2 points by lars 11y ago|1 comments
64.
by lars 11y ago
It's almost certainly the tide. I'm familiar with the area around Torghatten. The high tide/low tide difference there can be ~2 meters, more than enough to explain all the variation in sealevel.
65.
by lars 11y ago
There's a huge difference between a) not finding an effect, and b) finding that there is no effect. The former is the default result, and is what you get when your test lacks statistical power (from having a too low sample size). It su
66.
by lars 11y ago
That was exactly the problem, bottle+gevent serving static files. It's moved behind nginx now. (But you might have to wait for a DNS propagation before you get to the new server.)
67.
by lars 11y ago
Working on it:) It's getting a bit more traffic than expected at the moment.
68.
Auto-Generating Clickbait with Recurrent Neural Networks(larseidnes.com)
251 points by lars 11y ago|64 comments
69.
by lars 11y ago
This is the first I'm hearing of incremental-dom. Can anyone explain what it is? According to their readme: > It differs from the established virtual DOM approach in that no intermediate tree is created (the existing tree is mutated
70.
by lars 11y ago
With SGD, you know how to take steps in the direction of a better solution. You don't have that with genetic programming, there you just know if a solution happened to be better.
71.
by lars 11y ago
How about US6145506 A: Decorative penile wrap http://www.google.com/patents/US6145506
72.
by lars 11y ago
https://www.reddit.com/r/Bitcoin/comments/3fe92x/im_ashley_b... > So you're saying that the mt gox bank account was his own personal account? I wired money to his personal account?? > Yes Mayb
73.
by lars 11y ago
These are recurrent neural networks, not recursive. That's a different model, also used for NLP, but not so readily applicable for generating text AFAIK.
74.
by lars 11y ago
> getting very close to better-than-human performance I think they've all surpassed the human-level baseline by now. There's some stuff in there, like telling different dog breeds apart, that humans are not very good at.
75.
by lars 11y ago
Norway's prison population is not more homogeneous though, 34% of people in Norwegians prisons aren't even Norwegian citizens. In US prisons the number of foreigners is at < 3%.
76.
by lars 11y ago
You seem to be talking about average wealth and social mobility. This is different from inequality, which is about the distance from the bottom to the top. Some of these results are counter intuitive. For example: There's data to show
77.
by lars 11y ago
The drug war is part of it. But the big thing to fix is income inequality. High income inequality is correlated with higher homicide rate, higher robbery rate, lower civic participation, higher mortality, lower social cohesion and lower wel
78.
by lars 11y ago
At least with iTunes for Windows, the experience is pretty horrid. I've found iTunes to be extremely hard to navigate, and generally difficult to use. Aesthetically, the app looks like a mix between an OSX app and a website, very out o
79.
by lars 11y ago
Both Theano and Caffe (dont know about Torch) will already use cuDNN if its installed.
80.
by lars 11y ago
This, combined with the fact that you can increment strings gives some 'interesting' results: $a = "2d9"; $a++; echo $a . "\n"; $a++; echo $a . "\n"; Output 2e0
81.
by lars 11y ago
I just linked my multi page blog post describing the problems I ran into.
82.
by lars 11y ago
FWIW, my tone in that article is exactly the tone I would have if you got me started on this in a real life conversation. If I sound like a blowhard, at least I'm an authentic blowhard, it's not some act I put on for the clicks. T
83.
by lars 11y ago
I got tons of both hate and love for that blog post. One of the few criticisms that I can accept is the "The problems are real, but don't matter that much." view. I don't agree, but I suppose people can have different ex
84.
by lars 11y ago
This looks like a good resource. But number 28 is wrong: > Always have a ‘.’ in your ng-models which insures prototypal inheritance. That isn't what that does. The egghead.io Angular videos also get this wrong. I really think that t
85.
by lars 12y ago
I'm not a logician by a long shot, so I probably can't explain that correctly. I think Gödel found a way to make a logical proposition refer to itself, and then found a way to assert provability. He could then construct the senten
86.
by lars 12y ago
That isn't quite right. With a sufficiently powerful formal system, you're forced to either have inconsistency or incompleteness - you're describing a system that is inconsistent. It's usually much better to have consist
87.
by lars 12y ago
We had AI systems that outperformed doctors in limited domains as early as in 1972 [1]. We've of course had massive improvements in machine learning in the 43 years since then. I think the practice of human doctors could be massively i
88.
by lars 12y ago
Disagree with most of this. You don't need a massive inventory for recommendations to be useful, and you don't need a massive data set of usage data to be able to do them. What really matters is how sparse the user x item matrix i
89.
by lars 12y ago
I don't see why anyone would want runtime type checking the way AtScript does it. With static typing, the compiler can catch type errors early, at the cost of rejecting some otherwise valid programs. With dynamic typing, that type chec
90.
Quantity Queries for CSS(alistapart.com)
82 points by lars 12y ago|25 comments

More ›