Re: Small Numbers

1

It's astounding how much we focus on this canonical idea of the real worker.

Sometimes the Law of Small Numbers is the Poisson distribution, the distribution of how many fish you catch (assuming the fish catching events are independent).


Posted by: dalriata | Link to this comment | 04- 2-17 4:15 PM
horizontal rule
2
Crom is strong! If I die, I have to go before him, and he will ask me, 'What is the law of small numbers?" If I don't know it, he will cast me out of Valhalla and laugh at me.

I did know the law of small numbers, but this particular formulation really sounds like it could come out of a humorous (or not humorous) fantasy novel.


Posted by: nosflow | Link to this comment | 04- 2-17 6:29 PM
horizontal rule
3

How about "If I don't know it, he will have a Prussian army horse kick me in the head."


Posted by: Moby Hick | Link to this comment | 04- 2-17 6:39 PM
horizontal rule
4

It has wide application, outside mathematics as well as in. It will be proved by intimidation.


Posted by: JP Stormcrow | Link to this comment | 04- 2-17 7:01 PM
horizontal rule
5

Maybe there aren't a lot of small whole numbers, but there are a lot of numbers between 0 and 1, and real men use real numbers while loading coal into their testicles.


Posted by: fake accent | Link to this comment | 04- 2-17 7:27 PM
horizontal rule
6

The law of small numbers is that all phenomena are directly or indirectly related to the number five, and this relationship can always be demonstrated given enough ingenuity on the part of the demonstrator.


Posted by: Todd | Link to this comment | 04- 2-17 7:51 PM
horizontal rule
7

I had some far sighted friends back in the 1970s. They worked for a group at the MIT AI lab dealing with Very Small Databases (VSDB). They even had a bunch of such databases decorating their office. Now, it sounds like the antidote to big data.

P.S. I though that mathematicians consider small numbers those that can be written in finite time.


Posted by: Kaleberg | Link to this comment | 04- 2-17 8:57 PM
horizontal rule
8

I love the idea of Small Data as the new trend.


Posted by: ajay | Link to this comment | 04- 2-17 11:42 PM
horizontal rule
9

Artisanal data.


Posted by: teofilo | Link to this comment | 04- 2-17 11:42 PM
horizontal rule
10

"It's astounding how much we focus on this canonical idea of the real worker."

And also the related idea of the Real Job (you know, as in "he's a career politician, he's never had a".) Funnily enough Real Jobs generally tend to be those done by right wing men.


Posted by: ajay | Link to this comment | 04- 2-17 11:44 PM
horizontal rule
11

If I'd thought of it in time I would have definitely done an April Fool article about the "Small Data" trend. Using not-very-powerful software to draw conclusions from, say, four or five pieces of data.


Posted by: ajay | Link to this comment | 04- 3-17 1:36 AM
horizontal rule
12

Oh, I've heard "small data" used to mean selecting the most informative data set and doing a project that fits on your laptop rather than loading petabytes of stuff into some huge cloud and hoping the ML gods smile on you.


Posted by: Alex | Link to this comment | 04- 3-17 2:53 AM
horizontal rule
13

Also the Tversky/Kahneman paper "Belief in the Law of Small Numbers" (believing incorrectly that short random sequences have the same properties as long sequences)


Posted by: Kreskin | Link to this comment | 04- 3-17 4:37 AM
horizontal rule
14

I love the idea of Small Data as the new trend.

Artisinal, minimalist data.


Posted by: heebie-geebie | Link to this comment | 04- 3-17 5:30 AM
horizontal rule
15

Data borrowed from your parents, because you don't pwn any yourself.


Posted by: Mossy Character | Link to this comment | 04- 3-17 5:33 AM
horizontal rule
16

It's not an anecdote, it's Tiny Data. #tinydata


Posted by: ajay | Link to this comment | 04- 3-17 5:35 AM
horizontal rule
17

Trying to draw meaningful conclusions from a limited amount of crap data is like, 20% of my job.


Posted by: Spike | Link to this comment | 04- 3-17 6:40 AM
horizontal rule
18

Sort of the same here. Data points cost thousands of dollars in some of the things we do.


Posted by: Moby Hick | Link to this comment | 04- 3-17 6:46 AM
horizontal rule
19

Get a nanodegree in nanodata science.


Posted by: fake accent | Link to this comment | 04- 3-17 6:57 AM
horizontal rule
20

That should totally be a thing.


Posted by: Spike | Link to this comment | 04- 3-17 7:10 AM
horizontal rule
21

Tying together two consecutive front-page posts, the nihilist Arby's twitter account is the second-best twitter (after dog_rates, and now that Dick Nixon isn't posting as much).


Posted by: Tom Scudder | Link to this comment | 04- 3-17 7:17 AM
horizontal rule
22

Someone just sent a link for a workshop on Extreme sources of data. I don't want to look up what that actually is, because I'll probably be disappointed.


Posted by: Spike | Link to this comment | 04- 3-17 9:27 AM
horizontal rule
23

There are over a million and a half college professors. Can we be real Americans now?


Posted by: Cala | Link to this comment | 04- 3-17 12:32 PM
horizontal rule
24

No. How many acres you are spread out over is what matters.


Posted by: JP Stormcrow | Link to this comment | 04- 3-17 12:36 PM
horizontal rule
25

You are all 3 acres of manure spread over all yourselves.


Posted by: heebie-geebie | Link to this comment | 04- 3-17 12:51 PM
horizontal rule
26

22: just be warned that the ICTP cafeteria food is not what you might hope for from Italian cuisine.


Posted by: essear | Link to this comment | 04- 3-17 1:01 PM
horizontal rule