Appliance energy consumption prediction and classification using federated learning

This repository uses federated learning (FL) to:
(i) train a model that can predict the energy consumption of appliances in a household, and
(ii) train another model that can classify the type of electricity appliances according to their energy consumption recordings.

The models are built using keras and aggregated with the tensorflow_federated package.

The repository is written in a more functional programming style to better suit integration into a jupyter notebook.

To run this repository on Google Colab, you can do so here!