Polynomial Regression and Parameter Optimization

[MSCSF]

Hi all,

I’d like to use the X-Partitioner noce together with a Polynomial Regression Learner and the Parameter Optimization Loop.
The workflow without the loop is working fine. Now, I want to connect the Parameter Optimization Loop Start and the X-Partitioner.

Usage of the flow variable:

The Parameter Optimization Loop Start should be connectes to the Numeric Scorer (after the X-Aggregator).

Why is’nt it possible to minimize the root squared error of the statistics node?

Many thanks

[elsamuel]

It is possible. You just need to configure the Numeric Scorer node to output the scores as flow variables.

[BerkayAkar]

@MSCSF

Why is’nt it possible to minimize the root squared error of the statistics node?

Many thanks

ın my opinioon you should try to table row to verable node

for example

[MSCSF]

Thank you @elsamuel and @BerkayAkar.
Why should I use either the flow variable output or the table row to veriable node?
What are the advantages / disadvantages?

Regards

[elsamuel]

The Parameter Optimization Loop End node requires a variable.

Therefore, if you want to minimize or maximize the output(s) from the Numeric Sorter node, then this must be converted to a variable.

Why try to do this with a separate node when the Numeric Scorer node itself can be set to output all the variables?

[BerkayAkar]

so fi you use the table row verable you can looking best parameter for all optimized parameters

[MSCSF]

Many thanks!
For me the “easy” solution using the flow variable putput worked absolutely fine

Thank you both