Theme Park Attendance Forecasting Pipeline with Python

Built a Python-based machine learning pipeline using MongoDB for data storage, integrating structured and unstructured data sources — weather, ride wait times, park schedules, and Reddit sentiment — to forecast Six Flags attendance for operational decision-making.

Combining heterogeneous data sources (numeric, time-series, and text sentiment) into a single forecasting pipeline required building custom ingestion and feature-extraction steps ahead of standard time-series modeling, reflecting real-world operational forecasting rather than a clean single-source dataset.