Re: [R] How to fix false positve rates?

From: Berton Gunter <gunter.berton_at_gene.com>
Date: Wed 15 Jun 2005 - 08:22:38 EST

 

-----Original Message-----
From: r-help-bounces@stat.math.ethz.ch
[mailto:r-help-bounces@stat.math.ethz.ch] On Behalf Of Anderson de Rezende Rocha
Sent: Tuesday, June 14, 2005 11:01 AM
To: r-help@stat.math.ethz.ch
Subject: [R] How to fix false positve rates?

Dear R-users,

I have a set of 12000 image samples. I can divide this set into two categories: training and testing. I need to classify the test set into a two qualitative outputs: true or false for some characteristic.

To do the classification I'm using the packages SVM in e1071 library and LDA in the MASS library. However, I'm with a great number of FALSE POSITIVE CASES in both classifiers, about 10/15%.

  1. How can I use these classifiers with a fixed 1% FALSE POSITIVE RATE?

Surely you jest! Classification is not hypothesis testing.You do not get to control either false positive or negative rates. You try to minimize (an unbiased estimate of) mislclassification rate.

2) Could anynone indicate me a non-linear SVM classifier R-package? 3) Could anynone indicate me an ADA-BOOST classifier R-package?

Look!
RSiteSearch('boosting', restrict = 'functions')

Cheers,
Bert Gunter

Thanks for all. I'm anxiously wait the answers. Best regards.



|Üñð뮧µñ| - Anderson de Rezende Rocha
Bacharel em Ciência da Computação - UFLA Mestrando em Ciência da Computação - UNICAMP Esteganografia e esteganálise digital
UNIVERSIDADE ESTADUAL DE CAMPINAS, SP - BRASIL < http://andersonrocha.cjb.net >

R-help@stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide!
http://www.R-project.org/posting-guide.html

R-help@stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html Received on Wed Jun 15 08:32:50 2005

This archive was generated by hypermail 2.1.8 : Fri 03 Mar 2006 - 03:32:38 EST