Dear R users, I'm new of the R, I really don't know much. I want classification some data (two class, many features and huge size of data) by using R. At this case, I want using Support Vector Machine, Bayes theory based classifier, Discriminant Analysis, Regression based at least. Which package should I using, and can I compare each classifier result by predictions? Thank you.
On Feb 11, 2011, at 1:36 PM, Jaeik Cho wrote:> Dear R users, > > I'm new of the R, I really don't know much. > > I want classification some data (two class, many features and huge > size of data) by using R. > > At this case, I want using Support Vector Machine, Bayes theory > based classifier, Discriminant Analysis, Regression based at least.http://cran.r-project.org/web/views/MachineLearning.html http://cran.r-project.org/web/views/Multivariate.html> > Which package should I using, and can I compare each classifier > result by predictions? > > Thank you. > ______________________________________________ > R-help at r-project.org mailing list > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide http://www.R-project.org/posting-guide.html > and provide commented, minimal, self-contained, reproducible code.David Winsemius, MD West Hartford, CT
> >> >> Which package should I using, and can I compare each classifier result by >> predictions? >>By prediction on the training data, emphastically no. By prediction on new data not used for training, yes. -- Bert
I mean, after done for the testing step, I want show which data classified to wrong class. That predictions. Jaeik Begin forwarded message:> From: Bert Gunter <gunter.berton at gene.com> > Date: February 11, 2011 3:00:47 PM CST > To: David Winsemius <dwinsemius at comcast.net> > Cc: Jaeik Cho <chojaeik at gmail.com>, r-help at r-project.org > Subject: Re: [R] About classification methods. > >> >>> >>> Which package should I using, and can I compare each classifier result by >>> predictions? >>> > > By prediction on the training data, emphastically no. By prediction on > new data not used for training, yes. > > -- Bert
On Feb 11, 2011, at 4:11 PM, Jaeik Cho wrote:> I mean, after done for the testing step, I want show which data > classified to wrong class. > That predictions.At this point my suggestion is that your (re?)-read the Posting Guide and determine whether you have adhered to the level of detail and specificity that is implied to be desirable or optimal for questions to r-help. There may be a language issue and without implying any moral issue, the provision of a worked example might be even more important here than it would be in a situation of a shared language. You might also consult the "How to ask good questions" link which IIRC is at the bottom of that document. (My apologies to Bert if this was a question that he really was hoping to answer.) -- David> > Jaeik > > > Begin forwarded message: > >> From: Bert Gunter <gunter.berton at gene.com> >> Date: February 11, 2011 3:00:47 PM CST >> To: David Winsemius <dwinsemius at comcast.net> >> Cc: Jaeik Cho <chojaeik at gmail.com>, r-help at r-project.org >> Subject: Re: [R] About classification methods. >> >>> >>>> >>>> Which package should I using, and can I compare each classifier >>>> result by >>>> predictions? >>>> >> >> By prediction on the training data, emphastically no. By prediction >> on >> new data not used for training, yes. >> >> -- Bert > > ______________________________________________ > R-help at r-project.org mailing list > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide http://www.R-project.org/posting-guide.html > and provide commented, minimal, self-contained, reproducible code.David Winsemius, MD West Hartford, CT