Hi, these algorithms aren't currently incorporated into RapidMiner. For getting a general impression how to work with filters, please go through all sample processes in the sample repository that comes with RapidMiner.
Greetings, Sebastian
PS: This is the free forum, the answers will not arrive after five minutes. Get used to it or buy support with guaranteed response time.
One more problem i faced in Rapidminer is when i use 'Forward Selection' and run the process it shows no data is delivered at 'port forward selection performance' and same when i use 'Logistic Regression' ´with the same data when i run it. It shows 'Learning scheme logistic regression does not have sufficient capabilities for the given data set, polynominal label not supported'
so please tell me where i am making mistake and how can i overcome this....
i was read the example so as i cross 15/26 there was some error to install WEKA pluginns....can you explain it to me what is it and how can i solve it?
Hi, to your first question: What is unclear on this messages? I guess your process design is wrong and does not deliver a performance to the respective port. Insert a cross validation with the needed child operators to estimate the performance. Logistic Regression doesn't support more than two classes (=polynominal label). Where is the problem with that error message?
No, I can't tell you, why it fails on your side, while everybody else can download it. At least not if you don't describe in more detail what error is shown and what happens.
yeah i got it now. one thing i want to know that is there any details description in RM to select the best from the generate weights of the features? because each operator gives different weighting does evey operator has some desciption inside the RM?
Hi, please look at the Help tab to get some more informations about the operators. If you need more details than that, please refer to the respective paper which should be findable on google scholar.
Thanks for your reply. i want to do CFS so tell me which one is the right block for this 'Weight by Correlation' or 'Performance(CFS)' because both are present in Rapidminer.
when i use Performance(CFS) it gives only some numeric value as a result but doesnt specify any class. how do i know about what is the meaning of this value which Performance(CFS) provides.
Contributor II
Answers
can you tell me about how can i work in rapidminer with there filters
Wilcoxon Rank Sum
Plus L take away R
Correlation based Feature Selection (CFS)
Please somebody reply me about this!
these algorithms aren't currently incorporated into RapidMiner. For getting a general impression how to work with filters, please go through all sample processes in the sample repository that comes with RapidMiner.
Greetings,
Sebastian
PS:
This is the free forum, the answers will not arrive after five minutes. Get used to it or buy support with guaranteed response time.
One more problem i faced in Rapidminer is when i use 'Forward Selection' and run the process it shows no data is delivered at 'port forward selection performance'
and same when i use 'Logistic Regression' ´with the same data when i run it. It shows 'Learning scheme logistic regression does not have sufficient capabilities for the given data set, polynominal label not supported'
so please tell me where i am making mistake and how can i overcome this....
Regards
i was read the example so as i cross 15/26 there was some error to install WEKA pluginns....can you explain it to me what is it and how can i solve it?
to your first question: What is unclear on this messages? I guess your process design is wrong and does not deliver a performance to the respective port. Insert a cross validation with the needed child operators to estimate the performance.
Logistic Regression doesn't support more than two classes (=polynominal label). Where is the problem with that error message?
No, I can't tell you, why it fails on your side, while everybody else can download it. At least not if you don't describe in more detail what error is shown and what happens.
Greetings,
Sebastian
one thing i want to know that is there any details description in RM to select the best from the generate weights of the features?
because each operator gives different weighting does evey operator has some desciption inside the RM?
please look at the Help tab to get some more informations about the operators. If you need more details than that, please refer to the respective paper which should be findable on google scholar.
Greetings,
Sebastian
i want to do CFS so tell me which one is the right block for this 'Weight by Correlation' or 'Performance(CFS)' because both are present in Rapidminer.
when i use Performance(CFS) it gives only some numeric value as a result but doesnt specify any class.
how do i know about what is the meaning of this value which Performance(CFS) provides.
as I said: Take a look at the operator documentation. You can't overlook this blue help button, can you?
Greetings,
Sebastian