U77w41/Forcasting-and-Anomaly-Detection ? reverse-engineered prompt

Build me a notebook based project that forecasts Walmart sales and flags unusual sales patterns from the same data.

I want a clean workflow that loads the data, explores it, prepares features, trains a few regression based forecasting models, and compares them so I can see which one works best. Please include polynomial regression, multiple linear regression, lasso, ridge, and elastic net, plus a time series model like ARMA for forecasting. Then add anomaly detection so the app can point out sales values that look off, using something like KNN, DBSCAN, and an LSTM autoencoder approach.

Please make it easy to rerun from the notebooks, save results and models, and keep the data workflow reproducible. If anything needs current package guidance or best practices, look up the latest docs online. I’d also like simple plots and a short summary of the findings so it’s easy to understand what the models are doing.