maduchs/Network-Intrusion-Detection-using-Machine-Learning-on-CIC-IDS2017 ? reverse-engineered prompt

Build me a notebook based project for multiclass network intrusion detection using the CIC IDS2017 dataset.

I want two Jupyter notebooks, one for data preparation and one for modelling and evaluation. The first notebook should load the seven CSV files from the dataset, combine them, clean the data, do a little exploratory analysis, split the data into train and test sets, scale the features, and save the processed outputs so the second notebook can use them. The second notebook should load that processed data, train and compare at least a random forest and an XGBoost model, tune the XGBoost model, evaluate it on the held out test set, and save the final trained model along with the scaler, label encoder, and feature names.

Please also include a confusion matrix and per class results, and make it easy to rerun from top to bottom if I drop the dataset into a data folder. If you need to check anything, look up the current docs online.