maclaimi20/hospitality-demand-forecasting-rwanda

# Hospitality Consumable Demand Forecasting ## Overview This project predicts demand for hospitality consumables in Rwandan TVET institutions using machine learning and the following obejctives acheived: 1. Analyse historical consumption patterns of hospitality consumables across selected TVET institutions. 2. Identify key demand drivers (enrolment, timetable, seasonality, events) influencing consumable usage. 3. To develop and train multiple machine learning models for consumable demand prediction. 4. Propose a scalable decision-support framework for consumable procurement in Rwandan TVET institutions. ## Dataset The dataset includes: - num_students - num_sessions_per_week - course_type - avg_session_duration, - historical_consumption - Date - Type of consumable - Quantity used - Number of students - Events/season - budget_allocated ## Model Used - Linear Regression - Random Forest ## How to Run 1. Clone the repository 2. Install dependencies: pip install -r requirements.txt 3. Run the notebook Nirereumuhire_Marieclaire_Rwanda_FinalProject.ipynb ## Key Findings 1. Historical consumption is the strongest predictor of demand 2. Student enrolment significantly influences consumable usage 3. Seasonal variation impacts demand patterns # Next Steps - Pilot with real data from MINEDUC/Rwanda TVET Board ## Author Marie Claire Nirere Umuhire Rwanda