*** DECISION TREE HELP ***

liamevans

Hi everyone, rather new to RapidMiner and some help would be appreicted.

I've applied my data set, set up the decision tree algorithm and generated my decision tree. Unlike most decision trees which have say one or two labels (play ball, no play ball, etc...), for each case within the dataset, each outcome is different.

I want to be able to input numeric values at each of the attributes along each rule. So lets say a rule is:

IF car = blue and engine = fast THEN solution is CAR_Y

I want to apply values for "car = blue" lets say 0.5 and then "engine = fast" is 0.9, so if we add them up CAR_Y = 1.4.

i'm we have like 40 rules, each with different "Cars" but for example we may have "car = red" etc...

Does anyone know if this can be done in RapidMiner?

Thanks in advance.

Kind regards,

Liam

land

Hi Liam,

you are talking about an algorithm that simply isn't something I would understand on the term Decision Tree. This isn't possible with RapidMiner, but you might build your own algorithm in it. Please refer to the WhitePaper how to extend RapidMiner in our shop for more information.

Greetings,
  Sebastian