Keras Autoencoder for Fraud Detection - Deployment

This workflow applies a trained autoencoder model to detect fraudulent transactions.


This is a companion discussion topic for the original entry at https://kni.me/w/H-jksHkg2nk7PYbx

In the last part of the workflow: Your action is to send email in case of fraud (prediction = 1), but the “Table Row to Variable Node” takes only the first row from “Rule Engine” node!

I guess the suitable workflow could be the attached one in order to scan all rows. What do you think?

Thanks for your kind feedback.