Setting the whole matrix dataset as a label

Let's say you have two-time series datasets, the first dataset has five explanatory variables, and the other one is a dependent dataset with about 1000 variables, each representing an individual cell. So how do you set the whole dependent dataset as a label for prediction or regression analysis?
If not possible, is it possible to loop the whole dataset and perform predictions for each cell?

Find more posts tagged with

AI Studio

Labels

Linear Regression

Predictions + Scoring

Accepted answers

All comments

BalazsBaranyRM

Hi!

In RapidMiner, one attribute is the label. It is marked as such using the Set Role operator.

You can use Loop Attributes to apply the same process on the different attributes. I guess you would loop on the dependent dataset's attributes, select the current attribute (identified by the loop macro), and join it to the explanatory variables. Then you'll have a model for each dependent attribute which you can store in your repository using Store and apply later.

Regards,
Balázs

YYH

Hi @TeeH,

Take a look at the operator "Multi Label Modeling" for time series
https://docs.rapidminer.com/9.9/studio/operators/modeling/predictive/ensembles/multi_label_model_learner.html

Basically this is doing some auto-loop for different prediction target, usually defined for time series forecasting

Hope it helps!

Cheers,
YY