Anonymous user
Talk:Multiple regression: Difference between revisions
added a link and a request for clarification
m (fix accidental damage) |
(added a link and a request for clarification) |
||
Line 4:
OK, there's now implementations in two languages. It's not clear to me how this is different from the polynomial fitting task either, but I'm a completionist (for [[Tcl]]) and not a statistician... —[[User:Dkf|Donal Fellows]] 10:34, 9 July 2009 (UTC)
An explanation from the Lapack documentation may be helpful. [http://www.netlib.org/lapack/lug/node27.html]
The idea is that you want to model a set of empirical data
<math>\{(x_1,F(x_1))\dots (x_m,F(x_m))\}</math>
by fitting it to a function of the form
<math> F(x) = \sum_{i=1}^n\beta_i f_i(x)</math>
where you've already chosen the <math>f_i</math> functions
and only the <math>\beta</math>'s need to be determined.
The number of data points <math>m</math> generally will
exceed the number of functions in the model, <math>n</math>,
and the functions <math>f_i</math> can be anything, not just <math>x^i</math>.
I don't believe the Ruby and Tcl solutions on the page solve the general case of this problem
because they assume a polynomial model. I propose that the task be clarified
to stipulate that the inputs are two tables or matrices of numbers, one containing all values of
<math>f_j(x_i)</math> and the other containing all values of <math>F(x_i)</math>
with <math>i</math> ranging from 1 to <math>m</math> and <math>j</math> ranging from
1 to <math>n</math>, and <math>m>n</math>. --[[User:Sluggo|Sluggo]] 12:52, 9 August 2009 (UTC)
| |||