Logo Lanfrica

gre-eng/nigeria_disease_risk

Domain:

healthcare

Record type:

softwaremodel
Creator:
gre
Host:
AI-driven disease risk intelligence platform for Nigeria - cholera and Lassa fever surveillance, ML forecasting, and resource allocation # Nigeria Disease Risk Intelligence Platform An AI-powered web application for disease risk surveillance in Nigeria, focused on cholera and Lassa fever. The platform combines historical outbreak data, machine learning forecasting, and a risk-based resource allocation model into a single interactive dashboard. ## Live Demo (nigeria-disease-risk.onrend…) ## What This Project Does - **Cholera Hotspot Mapping** — ranks all 36 Nigerian states and the FCT by a composite risk score based on case burden, case fatality rate, and population-adjusted incidence. - **Lassa Fever Surveillance** — visualises national weekly case and death trends from NCDC situation reports (2020–2025). - **AI-Powered Forecasting** — trains and compares three machine learning models (Linear Regression, Polynomial Regression, Random Forest) to forecast Lassa fever cases 12 weeks ahead, automatically selecting the best-performing model by RMSE on a held-out test set. - **Model Validation** — transparently reports RMSE, MAE, and R² for each model, with a chronological 80/20 train-test split to avoid data leakage in time-series forecasting. - **Water Access vs Cholera Correlation** — overlays World Bank water access data (indicator `SH.H2O.BASW.ZS`) against cholera case trends to support WASH (Water, Sanitation and Hygiene) investment arguments. - **AI-Driven Resource Allocation** — proportionally allocates ORS packets, vaccines, and health workers across states based on computed risk scores. - **LGA-Level Ranking** — surfaces the top 10 highest-risk Local Government Areas for targeted intervention. ## Tech Stack - **Backend**: Python, Flask - **Machine Learning**: scikit-learn (Linear Regression, Polynomial Regression, Random Forest) - **Data Processing**: pandas, NumPy - **Frontend**: HTML, CSS, JavaScript, Chart.js (rendered via Flask/Jinja2 templating) - **Deployment**: Render (Gunicorn as the production WSGI server) - **Progressive Web App**: installable on phone, tablet, and d …