mjw

18 ms·

91.
by mjw 14y ago
The keyword syntax is transformed into a keyword instance by the reader, and keyword instances implement IFn as objects on the JVM, hence can be used as functions under clojure's evaluation rules. As I understand it, the roles which things
92.
by mjw 14y ago
Perhaps you could view this "A/A/B" test as a very crude form of http://en.wikipedia.org/wiki/Bootstrapping_(statistics) method? At least if you're resampling A1 and A2 from a pool A and then doing separate A1/B and A2/B tests and looking
93.
by mjw 14y ago
Perhaps the US constitution isn't the be-all-and-end-all of political thought, and people have valid disagreements on which parts of it are important?
94.
by mjw 14y ago
> When and if you need it, you can basically make your own type system, tailored to your application. When you do decide you need to think about and enforce types though, it helps a lot to have a clean, well-thought-out formally-speci
95.
by mjw 14y ago
In case the measure theory put anyone off: you don't need Dirichlet Processes for plain LDA, just the finite-dimensional http://en.wikipedia.org/wiki/Dirichlet_distribution (which isn't so bad and a very useful tool in Bayesian stats as t
96.
by mjw 14y ago
I've taken courses from people in a few of these overlapping but historically-different camps recently, e.g.: - Frequentist statisticians - Bayesian statisticians - Old-school AI researchers - Statistical learning theorists - Bayesian machi
97.
by mjw 14y ago
> Fragile, secretive and specific techniques ... parameters will be supplied to you for a price by the inventors-researchers' companies This seems a strange/dated/paranoid view of machine learning. Perhaps it has been true historically
98.
by mjw 14y ago
I ran into a nasty bug a while back which came down to a conflict between two JSON libraries in their definitions of #to_json on core classes.
99.
by mjw 14y ago
Nice one. Sounds like you guys have thought hard about the right balance between expressivity and allowing these structures to become excessively hard to reason about or traverse. Incidentally I found the use of topological manifold termino
100.
by mjw 14y ago
This is really interesting although a little dense/cryptic as slides. Does anyone have a link either to a slightly more leisurely exposition of the theory behind it, or perhaps a more user-focused summary of what kind of features and optimi
101.
by mjw 14y ago
Had it occurred to you it might be an off-the-cuff flippant example used as a bit of a joke? did 'penguin murdering' not give you a clue that this might not be a serious example intended to refer to someone's religion? British sense of humo
102.
by mjw 14y ago
True, although while the more practical mechanics of programming for data analysis can be learned easily, I wouldn't underestimate the time taken to really learn how to be scientific in thinking about and analysing data. I feel like the s
103.
by mjw 14y ago
Having worked with it for a few years, modelling music metadata is indeed an absolute nightmare. There are some efforts to standardise this stuff though, see http://www.ddex.net which a lot of the digital supply chain is starting to adopt
104.
by mjw 14y ago
Personally I thought the demographic here was supposed to be 'hackers'. I kinda wish the other various demographics would get a room: - Silicon valley gossip scenesters - The "I read an Ayn Rand novel, would you like me to turn this into a
105.
by mjw 14y ago
You probably guessed this but: promoters often have deals whereby they get a cut of the ticketing agency fees, say in exchange for exclusivity. So really the distinction presented to the consumer between fees and the face price of the ticke
106.
by mjw 14y ago
"Amex cards account for approximately 24% of the total dollar volume of credit card transactions in the US" Personally I use amex for all large transactions because they give me a cut (I think around 0.5% cash equivalent?) in air miles, but
107.
by mjw 14y ago
Don't have much to say about the relative merits of Kahn Academy specifically, but I think there's a balance to be had here. If you want to train people to be inventive in finding new ways to apply mathematics, and new mathematics to solve
108.
by mjw 14y ago
Although I don't buy the whole article, I think this is an interesting viewpoint. One could argue that the brand of libertarian politics often espoused in american hacker circles is pretty much the politics of this "meritocratic elite who c
109.
by mjw 14y ago
Combining the above with the original comment, check out Structure and Interpretation of Classical Mechanics: http://mitpress.mit.edu/sicm/ Physics explained by lispers! using consistent notation!
110.
by mjw 14y ago
Ruby and Python user here: If you ignore for a minute the complaints about the way Python modules are linked to the filesystem, I'd say there are actually some big advantages over Ruby in Python's module system. Mainly just that you actuall
111.
by mjw 14y ago
Multiple interpretations can also be useful purely as a pedagogical tool though. Using a range of metaphors to tell stories around some formal concept can make it easier to communicate and understand. That's not to say you should assume eve
112.
by mjw 14y ago
Ad-hoc as the use of a circuit solver might seem, there is some quite interesting machine learning literature on the use of resistive networks to do "graph transduction", e.g.: Mark Herbster, Resistive geometry for graph-based transduction
113.
by mjw 14y ago
What's the answer though? every politician should be hounded in every aspect of their private life just so we can avoid accusations of bias? That's going to drive good people out of politics. (Cue jokes about 'they already left a while ago'
114.
by mjw 14y ago
Is there a human filter on this stuff? I ask because, say a politician posts and then deletes something embarrassing on a personal level but which isn't particularly relevant to their political stance. They are human beings too and I don't
115.
by mjw 14y ago
Do you attempt to control for seasonal effects at all in your models? I'm wondering if allowing for a dependence on day-of-week, time-of-day and other relevant covariates, has any benefits over tracking fluctuations in a generic way?
116.
by mjw 14y ago
The useless answer is that they both do different things, so it depends what of those things you want :) One aspect of frequentist techniques that perhaps others haven't emphasised so much, is that they tend to give guarantees about expecte
117.
by mjw 14y ago
Very much agree. I'd also prefer not to encourage referring to people as "a Bayesian" or "a frequentist", as though it's a hard philosophical preference, a binary choice one has to make. I realise there are historical reasons for this, but
118.
by mjw 14y ago
See also http://www.mathworks.com/matlabcentral/fileexchange/19280-bi... amongst other density estimators, for a nice smooth plot of the estimated joint density. (Downside being there's still a somewhat-arbitrary choice to be made for the
119.
by mjw 14y ago
If you have enough point density for this to be a problem there's surely something better you can do though, whether it be binning the data or something a little more sophisticated like plotting a kernel density estimate.
120.
by mjw 14y ago
I'm not sure this works for me; I know what he means, but it's hard to view a rotationally-invariant object with positive curvature as being spikey. The analogy captures the 'what happens when you cut off a slice' property at the expense o

More ›