kylebgorman

8 ms·

31.
by kylebgorman 10y ago
Oh and of course, "gyspy"/"jitney" cabs/vans, which are quasi-legal $2 bus lines running bus routes in underserved communities.
32.
by kylebgorman 10y ago
So, for context, yellow taxis run "anywhere" but in practice, they are somewhat rare anywhere but lower Manhattan and at the airports. Green taxis, a new category, are only permitted to pick up hails outside of lower Manhattan. Th
33.
by kylebgorman 10y ago
Optimization is only a distant goal according to Guido---I think he said so in the PEP or the dev mailing list discussion thereof.
34.
by kylebgorman 10y ago
I feel the same way about C++ smart pointers, which are remarkably simple to use and understand.
35.
by kylebgorman 10y ago
I don't think we should incentivize moving to such places unless we plan to also (somehow) massively increase the opportunities available in such places.
36.
by kylebgorman 10y ago
I don't know how big "textfile.txt" is, but try it with a much larger body of text? Also you'll almost surely want a much better tokenization algorithm than that. Zipf's Law can only hold "at the limit" of
37.
by kylebgorman 10y ago
I don't think "amount of shared machine code" (DNA) is a good metric for functional similarity? What percentage of machine code do two small UNIX binaries (statically) linked against libc share? Potentially a lot. But obvious
38.
by kylebgorman 10y ago
Yet we have no clue what it's going to do next. Which is a serious indictment of the whole obsession with "brain mapping" and localization in neuroscience, IMO.
39.
by kylebgorman 11y ago
In Python, I'd use use a generator comprehension and str.title(). Assuming I haven't screwed up, then it's just: new_string = "".join(word.title() for word in long_string.split(" "))
40.
by kylebgorman 11y ago
There's no such thing as a PhD in deep learning. Your best bet is going to be doing machine learning at a top-tier CS school that has someone who is interested in it. But you're still going to have to take architecture, algorithms
41.
by kylebgorman 11y ago
The Economist has been open about their neoliberal beliefs since the magazine's inception.
42.
by kylebgorman 11y ago
I don't care about the presence of visual advertising, nor do I think cities would be nicer if it was all replaced with reproductions of Renaissance painting. It's not that I think I'm unaffected by advertising, it's jus
43.
by kylebgorman 11y ago
There are many groups working on non-invasive technologies that do the same using EEG. One commercial example, which I'm told is quite good: http://www.intendix.com
44.
by kylebgorman 11y ago
I'm guessing: "sharpness" of image, good saturation, presence of (smiling?) human faces, non-human mammals facing the camera, bare human skin (?) (I agree that'd be cool.)
45.
by kylebgorman 11y ago
From this document, the chance that a perl 5 script is syntactically valid (but semantically different) perl 6 is awfully low. Now if the errors that arose said that you might be trying to execute perl 5 as perl 6, that'd be cool ;)
46.
by kylebgorman 11y ago
I'm short Twitter. The fact of the matter is it that Twitter should never have become a multibillion-dollar company. There is no barrier to entry into this space---any competent web developer could make a non-scaling Twitter in an afte
47.
by kylebgorman 11y ago
Of all products, food probably shows the least regional variance, at least in the US. (Source: I've lived like everywhere.)
48.
by kylebgorman 11y ago
C++ has both checked casts and unsafe casts.
49.
by kylebgorman 11y ago
CRFs directly estimate the posterior/conditional model you care about (it tells you how to tag things), whereas a HMM estimates the joint model which you then use for inference. The general feeling is that it is actually easier to lear
50.
by kylebgorman 11y ago
This is a great implementation, thanks. For those of you who need this to be fast and achieve the theoretical bounds, you may want to do something like this using a finite-state transducer toolkit. Here's an example with OpenFst: http
51.
by kylebgorman 11y ago
Plenty of technology exists for adaptation to speaker and/or dialect, it's just not considered particularly sexy. To phrase it in economic terms, you can spend your time doing algorithmic work to improve recognition slightly for y
52.
by kylebgorman 11y ago
Just my personal opinion, but when you have an indexed copy of the whole web , a few million OCRed-but-not-corrected books from previous centuries added to your LM are not going to improve 2015 speech recognition quality.
53.
by kylebgorman 11y ago
Don't quote me on this but I think there may have been an example of this in the most recent Stroustrup C++ book?
54.
by kylebgorman 11y ago
I just can't agree that a simple, linear-time operation like "tokenizing words and basic n-gram models out of them" is a tedious problem like you seem to be implying, nor do I feel a solution to this very-solved problem is &q
55.
by kylebgorman 11y ago
While it's not quite what you want (as the specification language is not JSON---it's more like C declarations), "protobufs" allow you to generate serializers and de-serializers in a number of languages. http:/&#x2F
56.
by kylebgorman 11y ago
Spacy's implementation, assuming it's roughly equivalent with the one syllog1sm blogged about, just does a greedy incremental parse so it only produces one candidate parse. It is possible to do incremental dependency parsing with
57.
by kylebgorman 11y ago
The problem is that the translation of that sentence is very bad. It is totally unidiomatic in English. I would render it "...one inhabited by the Belgae, another inhabited by the Aquitani, and the third inhabited by those who in their
58.
by kylebgorman 11y ago
But, where did you read standard error during the teletype era? Was it printed to a separate tape?
59.
by kylebgorman 11y ago
word2vec isn't "deep" in the relevant sense. The both skipgram and CBOW forms have a single hidden layer.
60.
by kylebgorman 11y ago
Doesn't work on my vanilla Mac.

More ›