shihabmuhtasim/Prediction-of-APS-Failure-in-Scania-Trucks ? reverse-engineered prompt

Build me a Jupyter notebook project that predicts APS failures in Scania trucks from the provided CSV files.

I want the notebook to load the training and test data, clean up missing values, explore the class balance and sensor patterns, and then train a few classification models to figure out whether a failure is related to the air pressure system or not. Please include sensible preprocessing, feature selection or feature cleanup, and a clear comparison of model performance using accuracy, precision, recall, and a confusion matrix.

Make it easy to run from top to bottom in order, with short explanations in markdown so I can follow what each step is doing. If you need to look up current best practices for handling imbalanced binary classification in scikit learn, feel free to do that. I’d also like the final notebook to be polished enough that it reads like a complete student project, with the main results shown clearly at the end.