This KNIME Hub Space is dedicated to example workflows and additional files for the verified component “Counterfactual Explanation (Python)” available here: kni.me/c/wpVF3wtKLnH5V-IR In the folder “01_KNIME_Workflows” you can find the example workflows to explain predictions in KNIME from Keras and scikit-learn models. In the folder “02_Jupyter_Notebooks” you can find Python scripts to train and package models externally. Please note that you can also use KNIME to train models in Python. The file “custom_class_data_processing.py” defines the custom Python class used to normalize the data in training. If you use the “Counterfactual Explanation (Python)” component in a new workflow please add this Python file to the workflow folder via your file system. The data used for training the models was a sample taken from the 1994 Census database available at: archive.ics.uci.edu/ml/datasets/adult