Keras Autoencoder for Fraud Detection - Integrated Deployment

This workflow replicates the exercises of session 2, combining training and deployment using Integrated deployment. The purple boxes contain deployment nodes, that will be included in a new workflow Autoencoder_deployed. You can call the deployed workflow to get a prediction. The model and parameters will be up-to-date with the training and won't be executed at each call.

External resources

• L4-DL course slides