I have been using KNIME since past a month now and am new here. In my project I am trying to predict multiple targets and in python’s sklearn library there are two ways to do this:
- Single Target method (simple)
- Regressor Chain method
Each of the method makes single model for each target and then trains in different ways, but at the end of the day you are making different models.
We can do that too in knime but that process of making different models per target is manual.
My Question is that: Does KNIME support multi output feature?
If yes, how to do that? Which node to use?
Hi @aayushsmarten and welcome to the KNIME forum
in the KNIME Regression Learner nodes you can only provide one column as target. To automate the manual process of training many different models with the same input features and different target columns you can implement a loop.
Are you already familiar with the concept of flow variables and loops in KNIME Analytics Platform?
Umm, I am not familiar to that part actually yet. But sure that’s a way to get around with this problem. Thank you @Kathrin for your response!
I created a small example workflow for you, which uses flow variables. : Multi Target Regression – KNIME Hub
To understand the workflow I recommend to learn about flow variables. They are covered in our free L2 self paced courses which you can find here: Self-Paced Courses List | KNIME
Or if you prefer to read instead of watching videos you can learn more about flow variables and loops here: KNIME Flow Control Guide
This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.