Hongwei-Z/Federated-Learning-IoT-IDS ? reverse-engineered prompt

Build me a small research style app for a federated learning based intrusion detection system for smart home devices.

I want to be able to take the CIC IoT DIAD 2024 and CIC IoMT 2024 datasets, preprocess them for both binary and multi class classification, and then train and compare a few models in a federated setup and a centralized setup. The main goal is to detect anomalies and attacks in IoT and IoMT traffic, and to show clear results like accuracy, loss, precision, recall, F1, and confusion matrix style charts.

Please make it easy to run from notebooks or a main Python file, with clean data preparation, training, evaluation, and saved outputs in a results folder. If you need current package or dataset details, look them up online first. Keep the code readable so I can tweak it later and rerun experiments on my own machine.