aoki

19 ms·

181.
by aoki 9y ago
if your interests tend toward machine learning, you may run into calculus again. in particular, if you keep going past the "train a neural network" phase (for which very basic calculus is fine) to the optimization problems that ar
182.
by aoki 9y ago
it is (a standard HW problem). for bonus points, there's a connection between order statistics of the uniform distribution (such as max) and the beta distribution, of which the solution above is an example.
183.
by aoki 9y ago
i know what you mean. on the other hand, many of the problems in CS 229 are proofs and derivations that are very similar to those in the lecture notes (forcing you to understand the lecture notes in detail). as such, they can't just &q
184.
by aoki 9y ago
the SEE materials are from 2007.
185.
by aoki 9y ago
i suspect that what he's referring to is that he's heuristically minimizing a somewhat arbitrary (loss) function in a million-ish dimensions using the simple variants of gradient descent that work under these conditions. it sounds
186.
by aoki 9y ago
The Navy has been using such routing for decades.[0] Seems like the big win for Fleet Numerical would be in not having to run the workflow by hand [1] rather than adding (largely uninterpretable) neural networks to the problem... [0] http:
187.
by aoki 10y ago
am i the only one who thought: "hey, apple finally implemented the ANY key!" https://en.wikipedia.org/wiki/Any_key
188.
by aoki 10y ago
to your point about analysis, the folks with a CS education here (myself included) should ponder this quote from the intro: "...we have written this book to cover the theory likely to be useful in the next 40 years, just as an understa
189.
by aoki 10y ago
might be worth noting that "calculus by spivak" is famously mistitled and is what people today call an intro analysis text. (says so right in the intro :-) there are now "warm up" books (alcock) as well as even more basi
190.
by aoki 10y ago
ah, thank you. i did not really understand why this was really any more "surprising" than the 2D analogue (finite area under an infinite curve). there are at least some other people who think it's not :-)
191.
by aoki 10y ago
apparently had a preferred vendor in mind: http://www.bigassfans.com/
192.
by aoki 10y ago
click on the december 11 lecture.
193.
by aoki 10y ago
not sure i follow this. said man of peace was a navy officer and the article notes that he died wearing memorabilia from his service. do you have some reason to believe he would have agreed with you?
194.
by aoki 10y ago
ship naming has been random and political since forever. submariners used to call SSN 685 "lipscomb fish"...glenard p. lipscomb had nothing to do with the navy and did nothing but die as a congressman the year the ship was being n
195.
by aoki 10y ago
(1) removing the word "secret" from the HN title removes the only "interesting" element, since the fact that the stark was attacked by an iraqi jet is a matter of record. the article's claim is that it was a dassau
196.
by aoki 10y ago
well that's a blast from the past, thanks. though it did totally jump the shark when any bad news at any company of any size got to be "newsworthy."
197.
by aoki 10y ago
http://sfbay.wr.usgs.gov/sediment/southsfbay/depth_maps.html there is a marina in redwood city, but notice how shallow the bay is south of SFO (aside from the channel) - much of that is mudflats (lakes don't
198.
by aoki 10y ago
personally, i can't wait to see what the Silicon Valley (HBO) credits animation does with this next season. buildings standing up and trading places?
199.
by aoki 10y ago
MSR BARC was in downtown SF. did you mean the buildings around the late MSR-SV? no sign of that yet.
200.
by aoki 10y ago
i don't think that's quite right. when i was at intel the chief of staff role was the "technical assistant" or TA (as opposed to the administrative assistant or AA/EA); some TAs were business types but many had adva
201.
by aoki 10y ago
consider danah's example in which external studies show that blacks are convicted for drug crimes at a rate disproportionate to the actual crime incidence rate. the fact that the easily-available training data (court stats) is unrepres
202.
by aoki 10y ago
i don't think you meant "bad sample size," i think you meant "non-representative sample" or just "bad sampling design." the parent referred to bias. the example you're giving is one where the sample i
203.
by aoki 10y ago
if it makes you feel better, i don't think you were supposed to be inspired. "ML Ninja" is just the name of the rotation program. if your team sends you, it's because they need someone to get the training, not because th