headers
Kevin Chan | 4 Feb 2003 12:15
Picon
Favicon

JRip

Hi,

Where can I find the JRip classifier (Weka's implementation of the RIPPER 
rule learner)?

Thanks

_________________________________________________________________
Add photos to your e-mail with MSN 8. Get 2 months FREE*.  
http://join.msn.com/?page=features/featuredemail
Permalink | Reply | 1 comment |
headers
Cupidil | 4 Feb 2003 16:11
Picon
Favicon

Re: command line weka

Thanks Christian!

I've read the README file but it doesn't mention how
can I commit a test on a new data using a trained
data.

Thanks,
    Cupidil

--- Christian Schulz <ozric <at> web.de> wrote:
> i.e.
> java weka.associations.Apriori -t
> c:\weka\data\weather.nominal.arff 
> ...look at the Readme file for further info's.
> 
> regards,christian
> 
> ----- Original Message ----- 
> From: "No Name" <cupidil <at> yahoo.com>
> To: <wekalist <at> list.scms.waikato.ac.nz>
> Sent: Tuesday, January 21, 2003 2:19 PM
> Subject: [Wekalist] command line weka
> 
> 
> > Hi,
> > 
> > Does anyone know whether & how can I use weka as a
> > command line and not as a GUI in Windows?
> > 
> >
(Continue reading)

Permalink | Reply | 2 comments |
headers
Christian Schulz | 4 Feb 2003 16:31
Picon

Re: command line weka

Cupidil,

Store the model with the -d option
java weka.classifiers.j48.J48 -t c:\train.arff -d model1.mod
and then load it with the -l option

java weka.classifiers.j48.J48 -l model1.mod -T c:\test1.arff

The answer comes from gabi, January,8,2003 !
christian

----- Original Message ----- 
From: "Cupidil" <cupidil <at> yahoo.com>
To: "Christian Schulz" <ozric <at> web.de>; <wekalist <at> list.scms.waikato.ac.nz>
Sent: Tuesday, February 04, 2003 4:11 PM
Subject: Re: [Wekalist] command line weka

> Thanks Christian!
> 
> I've read the README file but it doesn't mention how
> can I commit a test on a new data using a trained
> data.
> 
> 
> Thanks,
>     Cupidil
> 
> --- Christian Schulz <ozric <at> web.de> wrote:
> > i.e.
> > java weka.associations.Apriori -t
(Continue reading)

Permalink | Reply | 1 comment |
headers
Hiroaki Yui | 4 Feb 2003 16:46
Picon

classifier

HI!

I am currently analzing the data, which consist of sales data: money, 
address, and so on. But I do not know whch classifier I should use  for
analyzing. If you know it, can you please give me an advice?

Regards
yui
Permalink | Reply | 1 comment |
headers
Richard Kirkby | 4 Feb 2003 21:32
Picon
Picon

Re: JRip

In the development version you'll find it as weka.classifiers.rules.JRip

Richard
--

Kevin Chan wrote:
> 
> Hi,
> 
> Where can I find the JRip classifier (Weka's implementation of the RIPPER
> rule learner)?
> 
> Thanks
> 
> _________________________________________________________________
> Add photos to your e-mail with MSN 8. Get 2 months FREE*.
> http://join.msn.com/?page=features/featuredemail
> 
> _______________________________________________
> Wekalist mailing list
> Wekalist <at> list.scms.waikato.ac.nz
> http://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
Permalink | Reply | 0 comments |
headers
Christian Schulz | 4 Feb 2003 21:54
Picon

Re: classifier

Hi,
it sounds like transaction-data for market-basket
analysis? and a common method are the 
association.rules and imho tertius is a interesting
"thing" for that .

christian

----- Original Message ----- 
From: "Hiroaki Yui" <hai44860 <at> par.odn.ne.jp>
To: <wekalist <at> list.scms.waikato.ac.nz>
Sent: Tuesday, February 04, 2003 4:46 PM
Subject: [Wekalist] classifier

> HI!
> 
> I am currently analzing the data, which consist of sales data: money, 
> address, and so on. But I do not know whch classifier I should use  for
> analyzing. If you know it, can you please give me an advice?
> 
> 
> Regards
> yui
> 
> 
> _______________________________________________
> Wekalist mailing list
> Wekalist <at> list.scms.waikato.ac.nz
> http://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
(Continue reading)

Permalink | Reply | 0 comments |
headers
Cupidil | 5 Feb 2003 12:15
Picon
Favicon

Re: command line weka

Thanks!

I cannot understand why when I train the classifier
using "java weka.classifiers.j48.J48 -t c:\train.arff
-d model1.mod" I receive predictions.
I cannot understand also why when I test on the
trained data (train.arff) I get only 96% correctly
which is the same accuracy that I received while
creating the model file.

I'd appreciate your help,

Cupidil

--- Christian Schulz <ozric <at> web.de> wrote:
> Cupidil,
> 
> Store the model with the -d option
> java weka.classifiers.j48.J48 -t c:\train.arff -d
> model1.mod
> and then load it with the -l option
> 
> java weka.classifiers.j48.J48 -l model1.mod -T
> c:\test1.arff
> 
> 
> The answer comes from gabi, January,8,2003 !
> christian
> 
>
(Continue reading)

Permalink | Reply | 0 comments |
headers
Cupidil | 5 Feb 2003 12:17
Picon
Favicon

Getting the Accuracy from Weka

Hi,

Is it possible to receive from Weka only the accuracy
of the correctly/incorrectly classified instances and
not the long output?

Thanks,
   Cupidil

__________________________________________________
Do you Yahoo!?
Yahoo! Mail Plus - Powerful. Affordable. Sign up now.
http://mailplus.yahoo.com
Permalink | Reply | 0 comments |
headers
jleone | 5 Feb 2003 13:24

Re: classifier


To me, it sounds like you're possibly trying classify customers based on
some criteria such as sales.  If that is the case and you're trying to
classify based on a binary dependent variable (e.g. "response") then there
are several methods.  J48, Logistic Regression and/or Neural Networks can
all provide good results.

If you are trying to predict a continuous value such as Sales, then an
M5Prime model tree could help.

Good luck.  I hope that helps.

"Christian Schulz" <ozric <at> web.de> <at> list.scms.waikato.ac.nz on 02/04/2003
03:54:31 PM

Sent by:  wekalist-admin <at> list.scms.waikato.ac.nz

To:   <hai44860 <at> par.odn.ne.jp>, <wekalist <at> list.scms.waikato.ac.nz>
cc:
Subject:  Re: [Wekalist] classifier

Hi,
it sounds like transaction-data for market-basket
analysis? and a common method are the
association.rules and imho tertius is a interesting
"thing" for that .

christian

----- Original Message -----
(Continue reading)

Permalink | Reply | 0 comments |
headers
mounir hamza | 5 Feb 2003 16:00
Picon
Favicon

CSVLoader Class

Hi everybody,

I'trying to use the CSVLoder (converter to the ARFF format) via the 
command line GUI but I am always obtaining exceptions.

I tried the following code :

java weka.core.converters.CSVLoder -file (I even tried without the 
dash)

Does any one has an idea about this problem.

Also how can we load the converted  file into another arff file

Thanks

Mounir
Permalink | Reply | 0 comments |

Gmane