kylebgorman

13 ms·

91.
by kylebgorman 12y ago
Not quite sure if this is what you want, but the noisy channel model (call it "Bayesian" if you must) spell checker I linked to above is inspired by the classic approach to speech recognition (and codebreaking, and machine transla
92.
by kylebgorman 12y ago
This is posted here about every three months. Norvig's approach assumes a very small fixed vocabulary (so it won't scale well to the basic multilingual plane), and can only recover errors with an edit distance of two or less. If y
93.
by kylebgorman 12y ago
Seconding the request for a date change. I remember reading this in 2003 when it was linked on /.
94.
by kylebgorman 12y ago
Your intuition that modern technology is slowing down language change, while entirely reasonable, is generally thought to be wrong by experts in language change. LETRA MAGNA - How do radio, television, films and popular entertainment affect
95.
by kylebgorman 12y ago
For something to not be "obviously false", I expect an intuitive mechanistic just-so story. So, how could fans, operating within standard tolerances, in closed rooms, possibly cause deaths? There is probably a just-so story that m
96.
by kylebgorman 12y ago
Consider using good Python style (like consistent use of whitespace) when trying to teach people to use Python (in any fashion).
97.
by kylebgorman 12y ago
The curriculum goes far beyond the knowledge of most IT professionals (not to mention our current graduate students), as it is focused on research; IT pros don't normally know much about, say, Python libraries for machine learning or s
98.
by kylebgorman 12y ago
Good idea, I'd love to enable online learning for this type of material. In the meantime, I'm putting handouts for the first unit ("UNIX") online, and the second unit ("Python") will use iPython notebooks. Ther
99.
by kylebgorman 12y ago
I am a jr. CS professor working in an applied field (computational linguistics). I have found myself repeatedly giving the same impromptu lectures to RAs and graduate students concerning various UNIX tools. This is a clear inefficiency (I c
100.
by kylebgorman 12y ago
Sign languages emerge when deaf kids (who aren't receiving other signed input) are together. To get the equivalent with spoken language, you'd have to put a bunch of (hearing) kids together on a desert island with no adult supervi
101.
by kylebgorman 12y ago
Peter Dodds, the leader of the team in question, isn't a computational linguist. I just read his CV and he has zero publications in CL journals or proceedings, and his Ph.D. is in mathematics.
102.
by kylebgorman 12y ago
If Julia catches on, it will mostly cannibalize R, not Python.
103.
by kylebgorman 12y ago
My hope is that Guido's plan to incorporate the type declaration module (`types`, a subcomponent of mypy) will encourage the Python team to clean up the issues brought up in TFA.
104.
by kylebgorman 12y ago
Why is this attributed to "Sweden" rather than to "Swedes" or "Swedish speakers"? * Swedish is the first language of about 5% of the population of Finland * Swedish is supposedly mutually comprehensible with No
105.
by kylebgorman 12y ago
While I'm not sure, I assume the author is referring to "error reduction". Error reduction is defined as def ER(accuracy_before, accuracy_after): delta = accuracy_after - accuracy_before return delta &#x
106.
by kylebgorman 12y ago
but "driverless car" does, as you would know if you had RTFA.
107.
by kylebgorman 12y ago
it represents the contents of the array to be sorted, I believe. x-axis is the index of in the array, y-axis is the value at that index.
108.
by kylebgorman 12y ago
Yes, we should ignore library code LOC as there's no associated cognitive/maintenance overhead, which is what we are really trying to count. I have happily used (C)Python for a decade without peaking at the source. Same goes for,
109.
by kylebgorman 13y ago
Is it possible that "taking all bomb threats deadly serious" is actually a poor policy? Arguments against: 1) base rate of airplane-related terrorism is very low 2) base rate of airplane-related terrorism that is preceded by a &qu
110.
by kylebgorman 13y ago
Were anti-vaxxers bound by the rules of valid inference, this would be a problem for anti-vax. Thimerosal, the compound supposedly responsible for the vaccine/autism link, has been phased out in the west but incidence rates continue to
111.
by kylebgorman 13y ago
I was quoting from earlier CDC reports. No study has looked at all 50 states to my knowledge (though tell me if you know of one), and even between participating states there are irregularities due to what kinds of records the CDC has access
112.
by kylebgorman 13y ago
Not a psychiatrist, just a researcher, but you can't get a diagnosis of autism spectrum disorder without a second cluster of symptoms related to repetitive behaviors and restricted interests. Shyness and speech/language problems n
113.
by kylebgorman 13y ago
What makes you think higher average IQ is associated with autism? (Rain Man doesn't count.) The state with the highest reported rate of autism spectrum disorder is Utah, followed by New Jersey. The states with the lowest reported rates
114.
by kylebgorman 13y ago
Autism spectrum disorder diagnosed using a standardized "observation" (the ADOS) and/or a highly structured parental interview (the ADI-R). There are peer-reviewed papers which report on the validity and reliability of these
115.
by kylebgorman 13y ago
Could you explain exactly why that triggered so much outrage for you? (As it turns out, this doesn't meet the definition of an ad hominem since there is no attempt to invalidate any arguments from "0pointer.net".)
116.
by kylebgorman 13y ago
This is an absurd idea. Enormous amounts of alcohol are consumed in airports and aboard planes, which impairs people's abilities to attend, act quickly, precisely, etc. If something bad happens, I'm going to be way more worried ab
117.
by kylebgorman 13y ago
I'd say more that it's approaching "domain-specific language". (I think R's syntax is just horrendous; love the libraries, though! ggplot2 included. Thanks Hadley!)
118.
by kylebgorman 13y ago
> some conflicts of interest are viewed as obvious. If I publish a journal article with my affiliation being Genentech, I don't have to state that my work is owned by Genentech. That's simply NOT true...Nature's guidelines
119.
by kylebgorman 13y ago
Nature should declare its conflict of interest...it requires article authors to do so.
120.
by kylebgorman 13y ago
Why are there so many bars in airports, then?

More ›