Polynomial Regression Model Evaluation


Are there any tools for evaluating a polynomial regression model, like computing R2 (R squared), F-statistic, and p-value, etc in KNIME?

Currently polynomial model only produces squared error for each row and a scatter plot.



Hi Larry, I beleive with Formulize/Eureqa you have similar options, and can give even more general results. There is a plugin to use it with Formulize. (Disclaimer: I am the plugin's developer.) With a  private server your data do not have to leave your computer.

(Eureqa was it's first name, then Formulize, now back to Eureqa. I have not yet switched back to Eureqa.)


Edit: Sorry, I just realized you were referring to the evaluation. Currently it is not returned for all kind of metrics available, so maybe this is not interesting for you.

Thanks, I will give it a try as it looks like an interesting tool to discover hidden relationships in the data.


I think, that realization of this features on base level - at KNIME regression nodes will be great. When we wright results in articles we must use this parameters and I compute it in spreadsheets, or using Math formula node.