> If I wished to use the meta AutoWEKA classifier, what changes, if any,
> would be required to WekaApp.data
I haven't used AutoWEKA myself. In theory, it should just act like any
other classifier (they all implement the same interface,
However, since AutoWEKA is a tool for parameter optimization (and can
take a long time, depending on the dataset), you would use that in an
offline fashion. You would then use the classifier/parameters that
AutoWEKA determined to be optimal for your data, build a classifier
with that and use this model in production.
Dept. of Computer Science
University of Waikato, NZ
+64 (7) 858-5174 (office)
+64 (7) 577-5304 (home office)
Dear Weka community,
I am writing a section on software comparison and I wish to describe Weka accurately, so I am turning to the community for help. I am a complete novice. I wish to build the simplest k-fold cross-validation workflow possible, preferably within a visual programming environment.
If I understand correctly, the visual programming environment would correspond to the KnowledgeFlow Environment?
Would the above image accurately (and not redundantly) represent a simple 10-fold cross-validation with Random Forest? The end goal is to observe the ROC curves, confusion matrix and evaluation scores.
Also, I am wondering whether Weka has any component/node recommendations? Either user- or rule-based?
Thank you so much for your input!
Hello, I downloaded Weka 3.8.5 for Mac OS Big Sur 11.6, which has java 8
(311). However, when I tried to launch Weka, I got the following error.
"Weka x-y-z is damaged and can't be installed. You should eject the disk
Please advise what to do to get Weka started. Thanks,
Letícia da Costa e Silva
Doutora em Meio Ambiente e Desenvolvimento - PPGMADE/UFPR
Pós-Doutoranda no Programa de Pós-Graduação em Agroecologia e
Desenvolvimento Rural Sustentável da Universidade Federal da Fronteira Sul
I have been ask by my security office to confirm if you software is affected by CVE-2021-44228 (Apache log4j) vulnerability. Can you please confirm if this is an issue with Weka.
IMPORTANT WARNING: The information in this message (and the documents attached to it, if any) is confidential and may be legally privileged. It is intended solely for the addressee. Access to this message by anyone else is unauthorized. If you are not the intended recipient, any disclosure, copying, distribution or any action taken, or omitted to be taken, in reliance on it is prohibited and may be unlawful. If you have received this message in error, please delete all electronic copies of this message (and the documents attached to it, if any), destroy any hard copies you may have created and notify me immediately by replying to this email. Thank you. Geisinger Health System utilizes an encryption process to safeguard Protected Health Information and other confidential data contained in external e-mail messages. If email is encrypted, the recipient will receive an e-mail instructing them to sign on to the Geisinger Health System Secure E-mail Message Center to retrieve the encrypted e-mail.
I tried to process a job in the experimenter by using the template "Spark: cross-validate two classifiers". When I start running the job, I alway get a failure message ", that something went wrong with the Shuffle Job. I'm using a Mac with intel processor. Any ideas what it is about?
Thank you in advance.
[ERROR] RandomlyShuffleDataSparkJob$2015179017|Job aborted due to stage failure: Task 1 in stage 3.0 failed 1 times, most recent failure: Lost task 1.0 in stage 3.0 (TID 9, localhost): java.lang.NoSuchMethodError: 'sun.misc.Cleaner sun.nio.ch.DirectBuffer.cleaner()'
java.base/java.lang.Thread.run(Unknown Source) "
I am running Experimenter with "Distribute" on multiple RemoteEngines.
I want to know if it is possible to use a number of RemoteEngines which is
larger than the total number of Runs.
For instance, suppose I do 10-fold cross validation and "From 1 To 10" runs
per data set and algorithm.
Suppose I have 10 algorithms and 10 data sets.
So if I understand, in total, there will be 10 runs (each of 10-fold cross
validation) for each of the 10x10=100 combinations (where a combination is
So 1,000 runs of 10-fold cross validation per run.
Now, suppose I have 25 cores (and I launched 25 RemoteEngines on the
From my experiments, it appears that only 10 remote engines are in use for
executing all the runs. It appears that the number of RemoteEngines in use
= the number of runs per combination. In the above, 10 runs per combination.
Is this correct?
It seems that it would be more beneficial if I could use all my cores, say,
25 cores, to run all the 1,000 runs, rather than use just 10 cores.
Is it possible to configure this?
Prof. Joel Ratsaby
Electrical and Electronics Engineering Dept.
I’m not sure you are aware that Weka classifiers can be wrapped with meta classifiers.
Bagging and AdaBoostM1 e.g.
These sort of chain the classifiers or create a hybrid classifier
My best Weka results are often Bagging RandomForest. Sort of odd since I found out later that RandomForest is sort of bagging but often true.
Does MLPlan support that sort of classifier was sort of the question.
Fwiw, I found at one point that you could even chain the meta classifiers…
I am using python-weka-wrapper and I need to specify the positive class for the JRip in the process of learning Ruleset. how can I specify the positive class when it has more instance than the other one. suppose for a dataset, after undersampling, the number of Majority class become less than the prime minority class but still I want to learn rules for the prime minority class( which now is the majority class after under sampling)
Thanks in advance
This email was Anti Virus checked by Security Gateway.