Values in the adaboost predictor

I am new to machine learning and i am currently using adaboost. After
creating the model, and testing it with the test data in the command
line , i got the predicted label followed by a value. Can any one help
me in knowing what those values are???

--

-- 
~~~~~~~~~~~~~~~~~~~~
Saravanan.V,
Ph.D Research scholor (DBT-BINC Fellow),
Centre for Bioinformatics,
School of Life sciences,
Pondicherry university, India
Puducherry-605 014.
brsaran <at> rediffmail.com
9884266881
~~~~~~~~~~~~~~~~~~~~~~~~~

_______________________________________________
Wekalist mailing list
Send posts to: Wekalist <at> list.scms.waikato.ac.nz
List info and subscription status: https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

help

Dear all, 

I am trying to compare a few algorithms against each other over some data sets. I have the exe files for
k-means and global k-means algorithms and when i run them i get the value of the objective function for each
of them (which makes it possible to compare them against each other). 

However i have problem with the WARD algorithm in WEKA, as in the output i do not see any value comparable to
other algorithms.  

The question is: firstly, is WEKA capable of comparing those mentioned algorithms against each other? In
other word does  WARD algorithm  give any output (e.g., objective function, time, etc.) comparable to
k-means variations? 

Secondly when i run WARD algorithm in WEKA i do not know how to interpret the results. Would you be able to give
me a source that could help me with the interpretation?

Cheers
Ken

_______________________________________________
Wekalist mailing list
Send posts to: Wekalist <at> list.scms.waikato.ac.nz
List info and subscription status: https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

Using SVM

_______________________________________________
Wekalist mailing list
Send posts to: Wekalist <at> list.scms.waikato.ac.nz
List info and subscription status: https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

Multiple lines in the Simple CLI

_______________________________________________
Wekalist mailing list
Send posts to: Wekalist <at> list.scms.waikato.ac.nz
List info and subscription status: https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

How is the “prediction” column in the Weka Naive Bayes output calculated?

Hi,

I did a Naive Bayes classification using 10 fold cross-validation
obtaining a table prediction on the test data that looks like this:

=== Predictions on test data ===

  inst#     actual  predicted        error  prediction         (name)
    1    3:no_chang 3:no_chang           0.943           (region_1)
    2    1:active_K 1:active_K                1                  (region_2)
    3    3:no_chang 3:no_chang           0.912           (region_3)
    4    3:no_chang 3:no_chang           0.858           (region_4)
    5    3:no_chang 2:active_G    +       0.518            (region_5)

I want to know how the "prediction" column is calculated. I know that
it goes from 0 to 1, 1 meaning that the prediction is "better", but
that's all I could find after a considerable amount of time googling
and browsing the Weka book.

I know there is plenty of information about Weka online, but I'm a bit
overwhelmed by it and can't easily find the answer to my simple
question. In addition to explaining how this is calculated, can
someone point me to a good detailed weka manual for a command line
user? The Weka book seems to focus too much in explaining how the GUI
work, which doesn't really interest me since I mainly work with the
command-line tools for the moment.

I Also posted the question here:
http://stackoverflow.com/questions/9516051/how-is-the-prediction-column-in-the-weka-naive-bayes-output-calculated

Thank you,

Juan

_______________________________________________
Wekalist mailing list
Send posts to: Wekalist <at> list.scms.waikato.ac.nz
List info and subscription status: https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

Re: Multiple lines in the Simple CLI

_______________________________________________
Wekalist mailing list
Send posts to: Wekalist <at> list.scms.waikato.ac.nz
List info and subscription status: https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

Re: Error Reading in ARFF File: Premature End of Line on Line 2

On 27/02/12 2:09 AM, fiona richards wrote:
> Hello,
>
> I'm getting a strange error when reading in a .arff file. Line 2 is (I'm
> assuming) the header of the file which as I used the sample program
> found on the wiki [http://weka.wikispaces.com/Creating+an+ARFF+file] I
> assume was correct.
>
> The file I'm trying to read in starts off as
>
>  <at> relation icat_results
>
>  <at> attribute Participant_ID numeric
>  <at> attribute Name string
>  <at> attribute Age numeric
> etc
>
> I'm not sure what is wrong with the format here, and why the Explorer
> interface is throwing the error.
>
> I'm be grateful for any insight into the problem, as I have yet to see
> an answer on Google for an error of this sort within the header,
> Many thanks,
> Fiona

I can't see any problem with the header you list above. Are you saying 
that the Explorer generates the error, or your own code?

Cheers,
Mark.

_______________________________________________
Wekalist mailing list
Send posts to: Wekalist <at> list.scms.waikato.ac.nz
List info and subscription status: https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

Re: Index of class attribute

On 27/02/2012, at 10:37 PM, Nicolas Martin wrote:

> Hello,
> 
> My arff file look like this : 
> 
>  <at> relation test
> 
>  <at> attribute fulltoken string
>  <at> attribute class {Y,N}
> 
>  <at> data
> {0 "information retrieval", 1 Y}
> {0 "harvesting information on the web", 1 Y}
> {0 "web mining and web retrieval", 1 N}
> 
> When I apply StringToWordVector filter, my arff file is transformed as : 
> 
>  <at> relation
'test-weka.filters.unsupervised.attribute.StringToWordVector-R1-W1000-prune-rate-1.0-T-I-N0-L-S-stemmerweka.core.stemmers.SnowballStemmer-M1-tokenizerweka.core.tokenizers.WordTokenizer
-delimiters \" \\r\\n\\t.,;:\\\'\\\"()?!\"'
> 
>  <at> attribute class {Y,N}
>  <at> attribute harvesting numeric
>  <at> attribute information numeric
>  <at> attribute retrieval numeric
>  <at> attribute web numeric
>  <at> attribute mining numeric
> 
>  <at> data
> {2 0.281047,3 0.281047}
> {1 0.7615,2 0.281047,4 0.281047}
> {0 N,3 0.281047,4 0.281047,5 0.7615}
> 
> As you can see, N class attribute is retained (in the third instance). In fact, I have to set my class
attribute to  <at> attribute class {DummyValue,Y,N} to get Y appearing in the transformed dataset : 
> 
>  <at> attribute class {DummyValue,Y,N}
>  <at> attribute harvesting numeric
>  <at> attribute information numeric
>  <at> attribute retrieval numeric
>  <at> attribute web numeric
>  <at> attribute mining numeric
> 
>  <at> data
> {0 Y,2 0.281047,3 0.281047}
> {0 Y,1 0.7615,2 0.281047,4 0.281047}
> {0 N,3 0.281047,4 0.281047,5 0.7615}
> 
> What is best way to proceed ?

You do not need a dummy class value. The data is in sparse format, which means that zero's are not explicitly
stored. This goes for nominal attributes as well as numeric. To convince yourself that all the
information is still there just load the data back into the Explorer and examine the it via the "Edit" button.

Cheers,
Mark.

_______________________________________________
Wekalist mailing list
Send posts to: Wekalist <at> list.scms.waikato.ac.nz
List info and subscription status: https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

Understanding the structure of Naive Bayes

_______________________________________________
Wekalist mailing list
Send posts to: Wekalist <at> list.scms.waikato.ac.nz
List info and subscription status: https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

Training and Testing

I am using a dataset of 7 nominal attributes for training. For testing the
same 7 nominal attributes but a different dataset with only some of the
nominal values repeats as training. It shows that training and testing not
compatible. I want to stick to the different testing set. What would be the
alternative?
Regards 
IQ
--

-- 
View this message in context: http://old.nabble.com/Training-and-Testing-tp33432174p33432174.html
Sent from the WEKA mailing list archive at Nabble.com.

_______________________________________________
Wekalist mailing list
Send posts to: Wekalist <at> list.scms.waikato.ac.nz
List info and subscription status: https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html