Store Sales Forecasting

- Developed a robust sales forecasting model using XGBoost to predict daily sales for a large retail chain

across multiple stores and product families, forecasting 15 days ahead.

- Engineered features such as lagged sales, trends, seasonality components to improve model performance.

- Optimized hyperparameters using Optuna to achieve a Root Mean Squared Log Error of 0.445 on the test

set