Sales Forecasting Using Time Series Data

• Data preprocessing to clean the dataset.
• Feature engineering to create time-series features (e.g., lag values, rolling means).
• Machine learning model training using XGBoost to forecast future sales.
• Evaluation of the model with metrics like Mean Absolute Error (MAE).
• Visualization of actual vs. predicted sales.

Dataset

The dataset is obtained from the UCI Machine Learning Repository. It contains retail transaction data from a UK-based online store.

• Link to Dataset: UCI Online Retail Dataset