Time Series Forecasting

Build a time series forecasting pipeline: data preprocessing with seasonal decomposition, feature engineering (lag features, rolling statistics, calendar features), model comparison (ARIMA, Prophet, LSTM, XGBoost), cross-validation with time series split, hyperparameter tuning, ensemble methods, and production deployment with retraining schedule