Re: Computerised authorship attribution
- From: "Phil Sherrod" <phil.sherrod@xxxxxxxxxxxxxxxxxxx>
- Date: Thu, 26 Jan 2006 19:50:02 GMT
On 26-Jan-2006, Matt B <mattb333@xxxxxxxxxxxxx> wrote:
> Phil - appreciate the offer, thanks.
>
> Unfortunately I'm supposed to do this whole thing myself. It's a major
> assignment and carries a weighting towards my final degree
> classification, so I can't use anyone else's code.
OK, that's fine. I was mainly trying to give you an example of a simple SVM
model that has been used successfully for the type of problem you have. The
example chart I referenced had only two dimensions with a splitting plane.
Of course greater accuracy can be achieved by using more dimensions, but
it's hard to visualize.
You might want to try to find the paper "Fung, Glenn. CS 525 Project.
Fall, 1998" and other articles that talk about classifying the authors of
the Federalist Papers.
> Besides which, I probably wouldn't be able to afford to pay you anyway!
I wasn't going to charge you anything.
--
Phil Sherrod
(phil.sherrod 'at' sandh.com)
http://www.dtreg.com (decision tree and SVM predictive modeling)
http://www.nlreg.com (nonlinear regression)
.
- References:
- Computerised authorship attribution
- From: Matt B
- Re: Computerised authorship attribution
- From: Ross Clement (Email address invalid - do not use)
- Re: Computerised authorship attribution
- From: Matt B
- Re: Computerised authorship attribution
- From: Phil Sherrod
- Computerised authorship attribution
- Prev by Date: Re: Computerised authorship attribution
- Next by Date: Re: estimating the inverse covariance matrix
- Previous by thread: Re: Computerised authorship attribution
- Next by thread: Re: Computerised authorship attribution
- Index(es):
Relevant Pages
|
|