Partial least squares regression/prediction


I'm working on a project for a industrial soft sensor with Knime. I have some data from sensors (temperature, ph, etc) and the variable to be predicted (Sulphur dioxide - SO2) is analysed in a laboratory each 2 hours.

My goal is to predict the SO2 each 10 minutes, I've read some papers about use PLSR (partial least squares prediction) for this case, but I didn't find this kind of predictor on Knime.

Can this work be done with Knime, or I have to develop something specific?





you can use a R node using the R PLS package.

Best regards

1 Like