xgboost on data set with both categorical and numerical data

I am working on diabetes data set with 35 attributes some are categorical like race, age groups drug dosages up down steady and no and some are numerical like time in hospital and admission id. and i would like to apply xgboost which is as I know works on numerical data only could you help with an example workflow with xgboost?

Hi @Monira_Mahmoud -

You can use XGboost for both classification and regression. The KNIME Hub has you covered:

In particular you might start with this relatively simple classification workflow:


@Monira_Mahmoud you could take a look at this example also including xgboost


I have not checked in KNIME but if XGB really only support numerical data (I would actually assume otherwise) you need to encode your categorical features (one to many node does one hot encoding)

1 Like

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.