Hitachi Vantara Pentaho Community Forums

Search Forums:

Type: Posts; User: Mike1961; Keyword(s):

Page 1 of 2 1 2

Search Forums: Search took 0.02 seconds.

  1. Replies
    1
    Views
    1,883

    Cant you just use Weka Explorer and under...

    Cant you just use Weka Explorer and under Preprocess-choose attribute- unsupervised-attribute- "NominaltoString" function?
  2. Thread: IBK Optimal K

    by Mike1961
    Replies
    1
    Views
    1,011

    Mark, Never mind on that, I figured it out....

    Mark,

    Never mind on that, I figured it out. But, it would be nice to have a toggle on the IBK KNN optimize to have it output the CV'd value too. :)
  3. Thread: IBK Optimal K

    by Mike1961
    Replies
    1
    Views
    1,011

    IBK Optimal K

    Hi Mark,

    I am trying to verify the K value found in IBK(crossvalidate = True and KNN set to max (number of instances) is correct.

    When using the crossvalidate IBK, I use that on "Use training...
  4. Replies
    1
    Views
    1,068

    Whichever model performs best on unseen data...

    Whichever model performs best on unseen data would be the right one. If I were to take a guess... the model on the left would be more robust because you used 8 inputs and the other
    has 10 inputs...
  5. Replies
    1
    Views
    1,363

    Package Installation Error

    Hi Mark,

    I am attempting to add packages on Weka-3-9. Some of the packages install with no error, but the ones that don't all have some "zip" error thwarting installation.
    Do you an idea to get...
  6. Replies
    1
    Views
    1,086

    I dont think there is a seed for IBK.

    I dont think there is a seed for IBK.
  7. Replies
    1
    Views
    1,055

    Artificial Time Index Explorer-Forecast

    Hi Mark,

    What is the meaning of artificial time index and the powers of time? If I have say 100 instances then would on day 100 the series would be:


    ArtificialTimeIndex*Lag_Diff-1 = 99 *...
  8. Solved it. For those experiencing slowdown like I...

    Solved it. For those experiencing slowdown like I have try looking at what's maxing out CPU usage.

    http://www.wintips.org/how-to-fix-svchost-exe-netsvcs-memory-leak-or-high-cpu-usage-problems/
  9. Replies
    4
    Views
    3,401

    I had a similar problem possibly with continuing...

    I had a similar problem possibly with continuing to upgrade on my laptop. My laptop was able to run 3.7.1 perfectly but when I upgraded to a later version 3.7.13 it couldn't even get the GUI to work....
  10. Replies
    3
    Views
    1,501

    Thanks for the offer but I don't write code. I'm...

    Thanks for the offer but I don't write code. I'm sure there are people on this forum who could though..
  11. Replies
    4
    Views
    1,697

    Ana, With ReliefF you might try removing some...

    Ana,

    With ReliefF you might try removing some variables(repeatedly) with RandomSubset(preprocessing filter) and see if ReliefF variable importance weight changes. I don't think ReliefF does a...
  12. Drastic Speed Slowdown -Old Files Need Deleted?

    Hi Mark,

    Are there any weka caches or old file dumping grounds that could be slowing my Weka bird from enjoying the brisk walking pace he enjoyed in his youth? :)

    Thanks,
    Mike
  13. Replies
    3
    Views
    1,501

    Recurrent Neural Network

    Hi Mark,

    I found a RNN package at : http://jsalatas.ictpro.gr/weka/


    Who's the one that adds this to Weka? Should I mention this as a project in JIRA if this can be done?


    Thank you,
    Mike
  14. deleted

    deleted
  15. Replies
    5
    Views
    2,229

    Woohoo!!! Success at last. :) Thank you Mark!!

    Woohoo!!! Success at last. :)

    Thank you Mark!!
  16. Replies
    5
    Views
    2,229

    After deleting all the packages it started back...

    After deleting all the packages it started back up.


    With all the packages installed, I get this log:


    2016-01-02 15:47:32 weka.gui.GUIChooser main
    INFO: Logging started
    [Weka]...
  17. Replies
    5
    Views
    2,229

    After deleting all the packages it started back...

    After deleting all the packages it started back up. Then I downloaded all the packages except openml and then failure upon trying to restart Weka. :(
  18. Replies
    5
    Views
    2,229

    Unable To Start Weka-3-7-13

    Hi Mark,

    I downloaded and installed the above package with JRE. After installation I launched Weka successfully but many of the available packages weren't showing up in Explorer. CLASSPATH is set...
  19. How Would I Find Diverse Attributes For Unlabeled Data?

    Hi Mark,

    The attribute data would be numeric and since I dont know the correct answer it would be unlabeled. What I want to do is use the Wisdom of The Crowds idea:
    ...
  20. Replies
    0
    Views
    487

    Number of Distinct Cases

    Message deleted
  21. Replies
    1
    Views
    1,642

    Merit of Best Subset Found

    Hi Mark,

    When using wrapper subseteval is a lower merit model better than a higher merit model? Logically, I would say the higher merit is a better model. But when I add noise to a variable it...
  22. Replies
    2
    Views
    812

    Thanks. I wasnt sure what Experimenter was...

    Thanks. I wasnt sure what Experimenter was randomizing. I will just let Experimenter take 30 swipes over the data in one step. :cool:
  23. Replies
    2
    Views
    812

    CV Parameter Selection

    Hi Mark,

    I am using Experimenter to search for parameter values. I set the training at 100 percent because CV PS already is doing cross-validations internally.

    Do I get roughly the same results...
  24. Replies
    8
    Views
    2,103

    Just make sure all your lags are moved down in...

    Just make sure all your lags are moved down in the spreadsheet and output gets shifted up(which is the future).

    Another form of future leak is running certain transforms (the ones that need all...
  25. Replies
    8
    Views
    2,103

    What version of Weka are you using? I remember...

    What version of Weka are you using? I remember early versions to be very finicky about train/test sets. In version 3-7-11 it never does this to me anymore and you don't have to use...
  26. Replies
    8
    Views
    2,103

    Is this time series data where the attibutes or...

    Is this time series data where the attibutes or output is shifted? If so, I suspect you have a future leak in your attribute(s). Let's say you have ten data points in your test set.If you remove...
  27. Replies
    3
    Views
    1,907

    [QUOTE=Mike1961;393912] Mark, Never mind...

    Mark,

    Never mind I see what the problem is. I misunderstood this to mean (M 2 20 1): start at 2, end at 20 incrementing by 1. Since I told it one step, I got the lowest(starting) parm every time....
  28. Replies
    3
    Views
    1,907

    [QUOTE=Mark;393856]Hi Mike, As far as I can...

    [QUOTE=Mark;393856]Hi Mike,

    As far as I can see, CV parameter selection and the Classify panel of the Explorer use nearly the same process to when running a cross-validation. First the full data...
  29. Replies
    3
    Views
    1,907

    CV Parameter Selection

    Hi Mark,

    On CV parameter selection(-X 10 -S 1) if I run that on Use training set is that the same as running a Cross-validation 10 folds with the same seed? The reason I ask is I cant get Cv PS to...
  30. Replies
    2
    Views
    841

    Actually it was a MLP. I thought initially it was...

    Actually it was a MLP. I thought initially it was a sign of a spurious variable. I have probed around with other data sets and know that is not the case so I'm left scratching my head on this. :)
  31. Replies
    8
    Views
    2,103

    Getting an equation from the MLP would be...

    Getting an equation from the MLP would be difficult. Just make a separate test set and choose it under: Classify> Supplied test set. In your test set use ? in the output column.

    If you absolutely...
  32. Replies
    2
    Views
    841

    Cross Validation Results

    Hi Mark,

    I occasionally run across models where the cv results(10 fold or LOO) are better than the training results. It is rare(ish) maybe like 1 in a 1000 chance. In your experience what would...
  33. Replies
    1
    Views
    789

    Mark, never mind. I just re-installed and the...

    Mark, never mind. I just re-installed and the temporary crisis is over.(wipes sweat off eyebrows) :)
  34. Replies
    1
    Views
    789

    Partition Membership

    Hi Mark,

    I downloaded some new packages to 3.7.11 and somehow in the process partition membership is gone. I use the GUI and it was there before the download. The package manager doesnt show that...
  35. Replies
    3
    Views
    3,660

    Hi Narges, In package manager(Gui chooser-...

    Hi Narges,

    In package manager(Gui chooser- Tools) download the gridsearch. On my package manager in 3.7.11 it is about half the way down.

    Regards,
    Mike
  36. Thread: SMOTE Error

    by Mike1961
    Replies
    1
    Views
    1,591

    SMOTE Error

    Hi Mark,

    I get a SMOTE error saying, " Problem filtering instances: Comparison method violates its general contract!". I know it is not in the Weka code itself because it functions perfectly on a...
  37. Replies
    2
    Views
    1,130

    Thanks, at least I know it wasn't just me.:)

    Thanks, at least I know it wasn't just me.:)
  38. Thread: Java ML

    by Mike1961
    Replies
    1
    Views
    718

    Java ML

    Mark,

    Would it be possible to get Java ML into Weka?

    http://java-ml.sourceforge.net/content/java-machine-learning-library-0.1.7-released

    There's alot of functionality to be added to Weka by...
  39. Replies
    2
    Views
    1,130

    GP Attribute Generator

    Hi Mark,

    Version 3-7-9:
    I am unable to get GP to work thru the CLI for batch files. After the string I use -b -i Train.arff -o TrainGP.arff -r Test.arff -s TestGP.arff.

    Can you see if you also...
  40. Replies
    1
    Views
    691

    MultilayerPerceptronCS

    Hi Mark,

    Is the NN using the valFile that I provide as a method to stop over-training?

    Thanks,
    Mike
  41. Replies
    3
    Views
    656

    Yes, like this as I opened on my desktop. All...

    Yes, like this as I opened on my desktop. All other features seem to work fine..

    8840
  42. Replies
    3
    Views
    656

    Cant Get Meta To Open

    Hi Mark, I can't get version 3.7.5 or 3.7.6 to open the Meta[+] box in Explorer-classify. This seems specific to my laptop(running Win7-64 bit) as my desktops have no problem with this. Also, on this...
  43. Replies
    2
    Views
    1,423

    The sourceforge website does have a genetic...

    The sourceforge website does have a genetic programming add on. It is using a 3.4 version shell so some of the more recent weka features arent available but
    I think does what you are looking for.
    ...
  44. Replies
    2
    Views
    1,807

    You can also in the 3.7 versions make separate...

    You can also in the 3.7 versions make separate arff files for training, testing and validating. So, you can take your training data csv file(or the other formats weka supports) and make an arff file...
  45. Replies
    1
    Views
    794

    ClassifierSubsetEval

    Hi Mark,

    What is happening when I have a hold out file loaded and have the use training switched to true? Is the evaluator using the training set and then
    verifying which variables did the best...
  46. Replies
    1
    Views
    1,047

    Running Weka 3.7.5 on USB?

    Hi Mark,

    How can I use 3.7.5 entirely off a USB drive? I want all the packages available but these are downloaded to the hard drive which is OK normally if I'm on my own computer.


    Thanks,...
  47. Replies
    1
    Views
    981

    Can't Find Vote In 3.7.5 Package Manager

    Mark,

    Where is the meta-classifier-vote listed at? I want to see if my installed version is the latest.

    Thanks,
    Mike
  48. Replies
    3
    Views
    1,753

    How do I choose what attributes to display in the...

    How do I choose what attributes to display in the predictions? All I get now is instance# of the crossvalidation not the ID# itself and actual - predicted-error-prediction columns.

    Thanks..
  49. Replies
    3
    Views
    1,753

    Remove Instances

    Hi Mark,

    What I would like to do is remove instances in which the probability estimate is below a certain threshold. Can I do this? How do I find in Explorer-Classify-Output Predictions- which...
  50. Thread: Co-EM

    by Mike1961
    Replies
    1
    Views
    1,410

    Co-EM

    Mark,

    In your book, there is a brief touch on co-EM. I initially follow what is going on but lose it in the following sentences.

    Can you give a layout of the concept and how to do this in WEKA?...
Results 1 to 50 of 93
Page 1 of 2 1 2
Privacy Policy | Legal Notices | Safe Harbor Privacy Policy

Copyright © 2005 - 2019 Hitachi Vantara Corporation. All Rights Reserved.