I noticed that different people mean different things when talking about a "model". What is the common meaning in different groups? Should be a different word used to prevent confusion?

Noticed meanings:

  1. statistical model: a set of functions.
  2. source model: a probability distribution.
  3. model of a prediction suffix tree: the structure of the tree. The parameters are not part of the model.
This question is marked "community wiki".

asked Jun 05 '11 at 09:28

Ivo%20Danihelka's gravatar image

Ivo Danihelka
25051115

wikified Jun 11 '11 at 05:30

http://searchweight.com/wp-content/uploads/2012/02/claudia-schiffer-3.jpg

(Mar 05 '12 at 12:11) Viktor Simjanoski

3 Answers:

Not an answer, but a brilliant debate: Breiman, Statistical Modeling: The Two Cultures.

This answer is marked "community wiki".

answered Jun 12 '11 at 06:54

denis's gravatar image

denis
2812812

From topic-models mailing list, link

“A model which, given a certain choice of subset from or probability distribution over a finite, discrete set of things called topics, gives you back side information which allows significantly better compression of some subset of some given data than any other choice or no side information at all.”

This answer is marked "community wiki".

answered Jun 12 '11 at 02:56

Xolve's gravatar image

Xolve
31224

From Hand et al's Principles of Data Mining, section 1.3: "A model structure, as defined here, is a global summary of a data set" (e.g. a linear regression), as opposed to a pattern: "pattern structures make statements only about restricted regions of the space spanned by the variables" (e.g. a probabilistic rule: if X > x1 then P(y) > p1).

answered Jun 06 '11 at 00:40

Lucian%20Sasu's gravatar image

Lucian Sasu
513172634

edited Jun 06 '11 at 01:28

Your answer
toggle preview

powered by OSQA

User submitted content is under Creative Commons: Attribution - Share Alike; Other things copyright (C) 2010, MetaOptimize LLC.