ed-wait-time-optimizer

End-to-end pipeline for optimizing emergency department wait times

🏥 Hospital Patient Volume Forecasting

Forecasting daily patient visits for a private hospital using time series analysis to support operational planning and data-driven decision making.

---

📌 Business Objective

Private hospitals face fluctuating patient demand, which directly impacts:

• Staffing levels
• Resource utilization
• Patient waiting times

This project forecasts daily patient volume to help hospital management make proactive operational decisions.

---

📊 Data

• Simulated daily patient visits (2 years)
• Includes weekly and yearly seasonality patterns
• Designed to reflect real private hospital demand behavior

---

🧠 Methods & Models

• Exploratory Data Analysis (EDA)
• Feature Engineering (lags, rolling averages, calendar features)
• Time Series Forecasting:
  • ARIMA
  • Prophet

---

📈 Model Evaluation

Models were evaluated using:

• Mean Absolute Error (MAE)
• Root Mean Squared Error (RMSE)

Model	MAE	RMSE
ARIMA	12.4	15.8
Prophet	9.1	12.3

Prophet was selected as the preferred model due to its accuracy and interpretability.

---

🖥 Interactive Dashboard

A Streamlit dashboard allows users to:

• Select forecast horizon
• Visualize expected patient volume trends

---

🗄 Database Design

A SQL schema is included to demonstrate how forecasted data can be stored and integrated into hospital reporting systems.

---

🛠 Tools & Technologies

• Python (Pandas, NumPy, Matplotlib)
• Prophet, Statsmodels
• SQL
• Streamlit

---

🎯 Role Alignment

This project was designed for Data Analyst roles in private hospitals, emphasizing:

• Business understanding
• Interpretability
• Actionable insights

---

🚀 How to Run

pip install -r requirements.txt
python src/data_generation.py
streamlit run streamlit_app/app.py