Classficationproblem

Flixport

Hello RM family,

I have a classification problem. I have a column with different classes, this column I want to take into account when validating with all classes. For example, the problem with Support Vector Machine is that I can only consider two classes and the rest is basically ignored. After consulting my professor, I should use the mapping operator as result. My solution was to add a new column to my CSV file and set the relevant class from the column to true, then the rest would be false. So I should consider all classes with the validation. I hope this approach sounds okay to you so far.

But what bothers me now is that it only gives me the true false values as a result, I would have given the classes instead . Of course my label is the new column with true false, but is there an another solution?

BR

MartinLiebig

Hi @Flixport ,

why are you just adding true/false to the table? Why not "ClassA" and "NotClassA"?

Best,

martin

Flixport

Hey @mschmitz

I can do it, but what is the difference between True/False and ClassA/NotClassA? I think the output would be the same

MartinLiebig

Hi,

for RapidMiner? None. We both map it to 0-1 in the background. I guess i did not undestand your question?

BR,

Martin