decision tree query
I am new to rapidminer & I want to learn the usage of decision tree operator. In decision tree is it possible to get our own approach rather than getting a big tree. Example if there are 3 columns in a table X column with records(x1,x2,x3,......xn) ; Y(y1,y2,y3,.....yn) & Z(z1,z2,z3)
In decision tree analysis I am getting all the combinations of x,y,z. But, what if I want to get a specific combination and probability like what is the probability of getting z1 when x1 works with y1 or z2 when x1 works with y3 and so on. and also, how to generate probabilities. I am a learner so please guide me to get the required output.
Thank You.
Best Answer
-
Telcontar120 RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 1,635 Unicorn
Decision Tree is not going to do what you want here, since there is no way for force it to select every possible existing combination. If you are interested in computing these types of conditional probabilities, you can do so fairly easily by using the Aggregate operator and grouping by whatever combination of attributes interests you and tabulating the counts.
1