WEKA machine learning software discussion

headers Chan Ai Ling | 1 Apr 2004 03:49

_______________________________________________
Wekalist mailing list
Wekalist <at> list.scms.waikato.ac.nz
https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist

_______________________________________________
Wekalist mailing list
Wekalist <at> list.scms.waikato.ac.nz
https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist

Hello all,

We're trying to connect Weka Explorer to an Oracle database, We can connect Weka to an Access DB, and the
Access DB connect to Oracle. We can also get Mathlab to see Oracle. When we try to connect Weka to Oracle we
get the error: "Can't open database: NULL." We know it is connecting to the ORacle DB because it detects bad userID/passowrds.

1. Is the Java code that creates / translates the connection string available?
2. Does anyone have an example of the strings they put in the "Open DB" GUI for an Oracle connection?
3. Should I even be using the JDBC-ODBC bridge?
4. Any advice? (Besides switching from Oracle - customer driven decision)

Thank you!
Vern

----------system data---------------------
OS: Win2000P
Weka: V3.4
Oracle:" V9i      DB:snook.orl     TNS Service: nook     Table:vwoodward.depot_data2
JBDC: sun.jdbc.odbc.JdbcOdbcDriver
OBDC: whatever MS uses  DSN:testdata
-------------------------------------------

----------Weka output---------------------
D:\Weka-3-4>java -jar weka.jar
Loaded driver: sun.jdbc.odbc.JdbcOdbcDriver
Connecting to jdbc:odbc:testdata
GenericObjectEditor: Problem making backup object
java.io.NotSerializableException: sun.jdbc.odbc.JdbcOdbcConnection
Executing query: SELECT * from depot_table2
-------------------------------------------------

--------------DatabaseUtils.props-------------
jdbcDriver=sun.jdbc.odbc.JdbcOdbcDriver
jdbcURL=jdbc:odbc:testdata
------------------------------------------------------

---------Explorer GUI-----------
jdbc:odbc:testdata
SELECT * from depot_data2
--------------------------------------

The problems that exist in the world today cannot be solved by the level of thinking that created them.
-A.Einstein 

_______________________________________________
Wekalist mailing list
Wekalist <at> list.scms.waikato.ac.nz
https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist

when i try to use a classifier on an arff file with a string attributes, 
I get the following error:

[pwaltman <at> pwaltman weka]$ java weka.classifiers.trees.j48.J48 -t test.arff
Cannot handle string attributes!

the format of my file is:
 <at> relation lgNaive

 <at> attribute string_ID string
 <at> attribute percentA {1,2,3,4}
 <at> attribute percentC {1,2,3,4,5,6,7,8,9,10}
....  more attributes
 <at> attribute class {Collagen,Elastin,Silk}

 <at> data
NP_892013.1,2,7,4,4,6,5,8,4,9,5,13,15,9,5,4,3,3,2,4,3,2,2,4,4,5,5,4,7,1,3,9,10,1,12,1,10,1,1,2,2,3,7,8,1,7,7,2,1,3,1,3,18,23,7,5,6,6,8,2,4,Collagen
NP_038627.1,1,8,4,4,6,4,9,4,9,6,12,15,6,5,4,3,3,3,7,6,6,4,3,6,6,5,4,9,4,3,2,10,8,17,1,2,1,1,2,3,5,3,3,3,5,12,11,1,3,3,11,18,18,1,16,13,7,22,1,7,Collagen

I have also tried putting the string attribute in double quotes, but 
still get the same error.  Has something changed in weka's 
implementation?  The README file says that it's supposed to support 
string attributes as a means of identifying the examples.  I've done a 
seardh of the mailist archive, and see posts from people who have had 
similar problems, but can't find any posts explaining how to resolve the 
problem.

as for weka version, I'm using weka 3.4

thanks,

Peter Waltman

IMHO the Decsion-Tree' Algorithm's from Quinlan are similar. 

weka.classifiers.rules.part.PART
weka.classifiers.trees.j48.J48

But a more comparable recursive-partioning algorithm to Chaid or CART  you 
find in www.r-project.org  with rpart.

christian

Am Donnerstag, 1. April 2004 03:49 schrieb Chan Ai Ling:
> Hi,
>
>
>  
>
>
> I was wondering does anyone know which algorithm in WEKA is similar to the
> CHAID (CHI SQUARED AUTOMATIC INTERACTION DETECTION) algorithm in SPSS?
>
>
>  
>
>
> Thanks a lot,
>
>
> Ai Ling
>
>
>
> MSN 8 helps ELIMINATE E-MAIL VIRUSES. Get 2 months FREE*.

IMHO why you want classify with a ID what looks like for me?
If it isn't only a ID and have additional information recode it to nominal 
attribute.

christian

Am Freitag, 2. April 2004 21:23 schrieb Peter Waltman:
> when i try to use a classifier on an arff file with a string attributes,
> I get the following error:
>
> [pwaltman <at> pwaltman weka]$ java weka.classifiers.trees.j48.J48 -t test.arff
> Cannot handle string attributes!
>
> the format of my file is:
>  <at> relation lgNaive
>
>  <at> attribute string_ID string
>  <at> attribute percentA {1,2,3,4}
>  <at> attribute percentC {1,2,3,4,5,6,7,8,9,10}
> ....  more attributes
>  <at> attribute class {Collagen,Elastin,Silk}
>
>  <at> data
> NP_892013.1,2,7,4,4,6,5,8,4,9,5,13,15,9,5,4,3,3,2,4,3,2,2,4,4,5,5,4,7,1,3,9
>,10,1,12,1,10,1,1,2,2,3,7,8,1,7,7,2,1,3,1,3,18,23,7,5,6,6,8,2,4,Collagen
> NP_038627.1,1,8,4,4,6,4,9,4,9,6,12,15,6,5,4,3,3,3,7,6,6,4,3,6,6,5,4,9,4,3,2
>,10,8,17,1,2,1,1,2,3,5,3,3,3,5,12,11,1,3,3,11,18,18,1,16,13,7,22,1,7,Collage
>n
>
>
> I have also tried putting the string attribute in double quotes, but
> still get the same error.  Has something changed in weka's
> implementation?  The README file says that it's supposed to support
> string attributes as a means of identifying the examples.  I've done a
> seardh of the mailist archive, and see posts from people who have had
> similar problems, but can't find any posts explaining how to resolve the
> problem.
>
> as for weka version, I'm using weka 3.4
>
> thanks,
>
> Peter Waltman
>
>
>
>
>
> _______________________________________________
> Wekalist mailing list
> Wekalist <at> list.scms.waikato.ac.nz
> https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist

_______________________________________________
Wekalist mailing list
Wekalist <at> list.scms.waikato.ac.nz
https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist

Harry Wells wrote:

> Dear All,
> 
> I was wondering how might data mining be applicable to sales forecasting, for example shoes.

I am not sure if you can do forecasting using WEKA. Perhaps the WEKA 
team can clarify this for this list. As my understanding of the area of 
forecasting based on my experience in numerical computing , you can 
attack this problem from different methodologies.

- Use System Identifications techniques (such as ARMA - Autoregressive
   Moving Average)

- Use Kalman Filtering techniques

- Use Wavelets Analysis techniques

- Use Hybrids Soft Computing techniques such
   as ANFIS (Adaptive Neuro-Fuzzy Inference Systems) and
   CANFIS (Co-Active Neuro-Fuzzy Inference Systems).

- and many more other methods, ...

Now I have used all of the above using MATLAB. The system ID toolbox in 
MatLab is very sophisticated. ANFIS & CANFIS are available in the Neural 
Network & Fuzzy Logic toolbox. I have written some Java classes in 
System ID for time-series analysis including forecasting. I did include 
some wavelets classes to fine-tune System ID algorithms. I did develop 
this work for a computational finance work I was involved before , for 
the analysis and forecasting of the stock-price movement. Wavelets does 
a very good job of decomposing the frequency components of the 
time-series and made the job of the SystemID algorithm easier in 
identifying trends, shocks,  and good prediction (forecasting) to a good 
degree. I have not done any Java work in Kalman Filtering yet but it is 
still on my to do list. I am currently trying to combine FuzzyJ (Fuzzy 
Logic Java Toolkit API - a commercial tool) which is available to be 
downloaded from the internet and JOONE (Java Object Oriented Neural 
Engine) which is an open source project in Java. My aim is to develop a 
Neuro-Fuzzy package where ANFIS & CANFIS would be available as well. 
Neuro-Fuzzy computing (soft-computing in general) is very popular in 
data mining at the moment.

There was an industrial mathematics week in Janurary (from 26th  to 
30th) of 2004 that was held at Auckland University, New Zealand which I 
attended and there were some presentations on forecastings. Neuro-Fuzzy 
techniques was shown of how to predict weather patterns (time-series) 
for a local power company (Transpower New Zealand). It is vital for this 
company to be able to forecast in advance the power demand and the wind 
velocity for the optimal performance of operating the wind-farm 
generators. Another group of mathematicians also showed by using the 
same data that Kalman Filtering is also a robust method for forecasting. 
Another group also used ARMA to do forecasting using exactly the same 
Transpower data. A local government research institute shows of how 
using wavelets techniques to forecast the frequency of earthquakes and 
where it might likely to occur. Now enough of that.

The JDMAPI (Java Data Mining API) which is JSR-73 will implement 
forecasting modules more likely to be in the next version 2. This a 
comment made to me by the lead-spec of JSR-73 Mark Honick of Oracle. 
Mark has also invited me personally to join the expert group for JSR-73 
which is responsible for the development of JDMAPI. If I will join this 
group (I have not decided yet) my main aim is to push for:

  - multi variate statistical analysis sub-package
  - systems IDs algorithms
  - soft-computing hybrids (Neuro-Fuzzy, Neuro-Genetics, Fuzzy-Bayesian,
    etc,.. )
  - numerical computing techniques as digital filters, wavelets
    and Kalman filtering.
  - Rough Sets Analysis

I have developed a full Java statistical API (Univariate & Multivariate) 
for my own work, some wavelets modules, system IDs and may be this work 
could be used as a basis to be modified and included in JSR-73 work for 
upcoming work in version 2 if I ever join the expert group for JSR-73.

Finally , there may be a way to do forecasting in WEKA , but I will 
leave that verdict to the WEKA team to comment.

Cheers,
Sione. Palu.

Hi

I have a question about multiclass SVM (Weka's SMO classifier) problems.

Say there are K classes in the problem.  The first question is how does weka determine which of the K classes
"wins" (i.e. gets the #1 ranking) for a given test instance?  And how does weka determine this in the
all-pairs case and is that different from the one-against-the-rest case?

Ideally, what I would really like is to get back a ranked list from weka of the K classes ranked in order of how
likely the software thinks that an instance matches the various classes.

On a related question, is there an easy way to get the distance of a instance vector from the separating
hyperplane?  Is such a measure the way a that weka ranks the different classes?

Any help would be most appreciated.  
--

-- 
Thanks!

Ari Chanen 

Ph.D. Scholar
University of Sydney
Sydney Language Technology Research Group

188	May 2013
358	April 2013
305	March 2013
302	February 2013
216	January 2013
200	December 2012
174	November 2012
178	October 2012
121	September 2012
180	August 2012
147	July 2012
192	June 2012
228	May 2012
220	April 2012
335	March 2012
233	February 2012
224	January 2012
285	December 2011
284	November 2011
247	October 2011
247	September 2011
326	August 2011
275	July 2011
249	June 2011
320	May 2011
274	April 2011
461	March 2011
365	February 2011
300	January 2011
234	December 2010
229	November 2010
338	October 2010
289	September 2010
349	August 2010
271	July 2010
345	June 2010
410	May 2010
336	April 2010
302	March 2010
264	February 2010
397	January 2010
332	December 2009
365	November 2009
362	October 2009
436	September 2009
453	August 2009
317	July 2009
455	June 2009
527	May 2009
501	April 2009
525	March 2009
362	February 2009
336	January 2009
222	December 2008
338	November 2008
331	October 2008
216	September 2008
250	August 2008
325	July 2008
282	June 2008
219	May 2008
354	April 2008
338	March 2008
264	February 2008
262	January 2008
197	December 2007
214	November 2007
288	October 2007
245	September 2007
176	August 2007
311	July 2007
276	June 2007
363	May 2007
410	April 2007
302	March 2007
231	February 2007
236	January 2007
155	December 2006
262	November 2006
197	October 2006
199	September 2006
188	August 2006
207	July 2006
289	June 2006
315	May 2006
289	April 2006
330	March 2006
238	February 2006
198	January 2006
212	December 2005
206	November 2005
253	October 2005
280	September 2005
271	August 2005
157	July 2005
267	June 2005
221	May 2005
286	April 2005
182	March 2005
117	February 2005
111	January 2005
108	December 2004
106	November 2004
81	October 2004
56	September 2004
69	August 2004
101	July 2004
143	June 2004
96	May 2004
87	April 2004
71	March 2004
104	February 2004
96	January 2004
70	December 2003
143	November 2003
97	October 2003
80	September 2003
87	August 2003
99	July 2003
90	June 2003
100	May 2003
123	April 2003
122	March 2003
119	February 2003
94	January 2003
40	December 2002
137	November 2002
122	October 2002
81	September 2002
55	August 2002
61	July 2002
43	June 2002
70	May 2002
35	April 2002
3	January 2001
2	March 2000

CHAID algorithm

Ranked classification for multiclass SVM?

Oracle connection

Oracle connection

error when using arff files with string attributes

Re: CHAID algorithm

Re: error when using arff files with string attributes

a Retail Application with WEKA - Please share your thoughts

Re: Retail Application with WEKA - Please share your

Ranked classification for multiclass SVM?

Mon	Tue	Wed	Thu	Fri	Sat	Sun

			1	2	3	4
5	6	7	8	9	10	11
12	13	14	15	16	17	18
19	20	21	22	23	24	25
26	27	28	29	30