An interactive, self-contained HTML presentation of a workshop on Time Series Analysis. This presentation uses a case study focused on the relationship between air pollution (PM2.5) and Tuberculosis (TB) in Kampala, Uganda, to demonstrate a complete data analysis workflow from initial analysis to policy simulation and recommendation.
The presentation walks through the essential methods and applications of modern time series analysis in the public health sector, including:
- Context Setting: Understanding the public health challenge in Kampala concerning air quality and TB.
- Interrupted Time Series (ITS) Analysis: A quasi-experimental design to evaluate the impact of interventions.
- Seasonality & Decomposition: Methods for understanding and handling cyclical patterns in health data.
- Time Series Feature Engineering: Creating derived variables like lags, calendar effects, and rolling statistics to improve model performance.
- Forecasting Essentials: A review of forecasting approaches from naive and statistical methods (ARIMA, Exponential Smoothing) to machine learning models (Random Forest, XGBoost).
- Model Evaluation & Interpretation: Using metrics like MAE/RMSE and explainability techniques like SHAP and LIME to understand model behavior.
- Policy Intervention Simulation: Using models to forecast the potential health and economic impacts of different policy scenarios.
This presentation is a single, portable file with several modern features:
- Fully Self-Contained: All styling and scripts are embedded in the HTML file. No dependencies or internet connection required to view.
- Full-Screen Layout: Designed to fill the entire browser window for an immersive experience.
- Interactive Navigation:
- Clickable "Next" and "Previous" arrows.
- Clickable dot indicators at the bottom for quick navigation.
- Keyboard support (left and right arrow keys).
- Responsive Design: The content is centered and readable on a wide range of screen sizes.
Viewing the presentation is simple:
-
Download the repository:
- Clone the repository:
git clone <repository-url> - OR download the ZIP file and extract it.
- Clone the repository:
-
Open the file:
- Navigate to the project folder.
- Open the
presentation.htmlfile in any modern web browser (e.g., Google Chrome, Firefox, Microsoft Edge).
That's it! The presentation will load and you can navigate through the slides.
This project is licensed under the MIT License. See the LICENSE file for details.