You are working on a forecasting task where past values, seasonality, and recent trends all matter. The raw data is noisy, irregular in places, and the useful signal is spread across time.
How would you approach feature engineering for a time series forecasting problem?
You are working on a forecasting task where past values, seasonality, and recent trends all matter. The raw data is noisy, irregular in places, and the useful signal is spread across time.
How would you approach feature engineering for a time series forecasting problem?