How to split decision tree learner manually


I have a table with attribute like acclon, acclat, pedal, speed, driver and I want to divide it manually like first division start from speed attribute into two part according to the set values & next division should divide from acclon into part according to set values and so on...

Can anyone help me how these implementation possible in KNIME. I try it but it is only dividing with one attribute like speed & till end of tree it will be dividing by speed only.

Do I need to create a new node for manually splitting, then it is possible??? or it is possilbe in decision tree learner node???

Thanks a lot



You are probably looking for an interactive decision tree learner, which we currently don't have. However, if you are looking for a way to partition your data given a criteria (that you need to implement, i.e. in a Java Snippet), you might want to check out the delegation nodes (from the Ensemble Plug-in) and used them to recursively go through you spitted data. If you want to implement your own node, have a look at the KNIME SDK.

Thank you gabriel

I install the Ensemble plug-in,

but what kind of input i need to connect the delegation nodes. I mean right now I have a table with some column as string and number. but when i connect to delegation node then configuration click is inactive.

So may be i need some other input.

If you know about it then please update me asap.