iskander

58 ms·

571.
by iskander 14y ago
He doesn't! Each mapper trains a model using stochastic gradient descent on a subset of the data and they are then combined into an ensemble. A more sophisticated alternative is the Terascale Learning Algorithm ( http://www.slideshare.net/p
572.
by iskander 14y ago
I'm pretty sure that his/her negative rating for "killing Americans without trial" is due to our recent practice of killing Americans without trial: http://www.csmonitor.com/USA/Justice/2012/0718/Families-sue-...
573.
by iskander 14y ago
>But the sex itself, is it that much different than "real" sex? Are there really that many people having boring sex to make the amount of people coming close to and/or surpassing porn a statistical anomaly? Oy vey, you're constructing a
574.
by iskander 14y ago
That was a terrible quiz for several reasons. 1) Some of the questions simply don't make sense, i.e. "Multinational companies are unethically exploiting the plant genetic resources of developing countries." The what resources? 2) It compl
575.
by iskander 14y ago
Cool, thanks.
576.
by iskander 14y ago
I took a biblical archaeology class as an undergrad and both our teacher and every text we read were very cautious about making this link. There simply isn't very much evidence for who the 'PRW (apiru or whatever) were; we don't even know i
577.
by iskander 14y ago
>...priests had strict dietary laws and they knew that Cysticercosis caused by tapeworms from swine. > Ancient Israelites just adopted a lot of customs from Egyptian Priestly caste...Not surprising, considering Moses was raised as E
578.
by iskander 14y ago
>Also, I think Python syntax is much more pretty while Matlab is not even designed to be a real language. If you're only coding the core of an algorithm (rather than a full-featured library with lots of plumbing), and your logic fits na
579.
by iskander 14y ago
Some people really do seem to get a lot done in Lush, so I'm not discounting its utility, but the language is sort of a mess. I took Yann's class and gave up in frustration after a few homeworks. I was very happy working in Matlab and relie
580.
by iskander 14y ago
1) I'm sad to find your totally reasonable comment downvoted. Decline in HN standards, etc... 2) An alternative scenario: what if a car dealership simply mis-priced a car? They're holding a rare race car but the owner asked his know-nothing
581.
by iskander 14y ago
>The gamblers unlawfully took advantage of the Golden Nugget when they caught on to the pattern and increased their bets from as little as $10 to $5,000 I don't understand why a casino feels like it's legally entitled to a profit even w
582.
by iskander 14y ago
Every test has a false positive rate and if you scrutinize close enough you can often guess reasons/justifications for those false positives. These rationalizations really aren't so important though--- the crucial bit is understanding that
583.
by iskander 14y ago
For you to see the app, think "I don't want to hear about drone strikes", and then (maybe, hopefully) realize that you're actively filtering out an undeclared war being waged on your behalf.
584.
by iskander 14y ago
Another trick I've seen in biomedical data mining is synthesizing new samples from the under-represented class (using an algorithm like SMOTE).
585.
by iskander 14y ago
Anecdotally, I've met many well-educated Russians who left in the mid-90s. I think, though, this was mostly due to fear for personal safety and a sense that their country was in a unrecoverable tailspin.
586.
by iskander 14y ago
I wish the window for fixing a comment was a bit longer...
587.
by iskander 14y ago
I haven't seen this book before but I did take Prof. Mohri's class a few years ago. His material is great but be warned that is's biased heavily towards theory (proofs and bounds instead of the more common/practical "rules of thumb"). You c
588.
by iskander 14y ago
I'm not sure the term "superpower" is well defined enough to argue over who is and isn't one, but I do take issue with the idea that the USSR was influential only via its military. Many, if not most, revolutionary movements for quite a whil
589.
by iskander 14y ago
>Of course, it is all economic waste, since none of the skyscrapers are occupied. No forces of demand or supply, no consumption locally of the product (people can’t afford it). I'm not qualified to judge how sustainable the whole Chines
590.
by iskander 14y ago
Let me turn the question around. I have about 400GB of data (spread over ~3000 HDF files, each containing 42 compressed columns). The only way I can imagine using BigQuery on this data is if I convert it all to csv files (which will take a
591.
by iskander 14y ago
If I remember correctly, BigQuery only lets you import data via local csv files, uploaded one at a time. That makes importing data sets of relevant size quite a pain.
592.
by iskander 14y ago
They really diverge when it comes to more advanced language features. I make heavy use of functors in my code and my (possibly mistaken) understanding is that F# simply doesn't support them. Ditto for first-class modules, polymorphic varian
593.
by iskander 14y ago
Jesus, my research project once depended on this file. The day we deleted it from the repository was a very happy one.
594.
by iskander 14y ago
I don't know much about this google project. Can autonomous cars (i) drive on the highway (ii) drive at night (iii) drive in snow or rain?
595.
by iskander 14y ago
One nitpick: You can't "kill the power" on a market-making algorithm co-located with the exchange somewhere in New Jersey.
596.
by iskander 14y ago
They weren't taking long-term speculative positions, their short-term market making had a major bug in which instead of buying low and selling high it was doing the opposite (thus losing a small amount on each pair of trades). For more de
597.
by iskander 14y ago
You can get more technical details from nanex: http://www.nanex.net/aqck2/3522.html Essentially, they were buying high and selling low. Many times a second.
598.
by iskander 14y ago
I actually love weave but I felt it was irresponsible to advertise its use without a huge warning.
599.
by iskander 14y ago
I've been doing data analysis in Python for the past 2ish years and I think it's a great choice. Before Python I worked in Matlab, which simplifies matrix operations at the expense of making everything else terrible. The main other contende
600.
Expensive lessons in Python performance tuning(blog.explainmydata.com)
163 points by iskander 14y ago|12 comments

More ›