Logo Lanfrica

mykiie85/AI-Driven-Vector-Control-Optimization-Model-for-Tanzania

Domaine:

healthcare

Type de record:

softwaremodel
Créateur:
myk
Hôte:
VCOM-Tz is an AI-driven vector control optimization platform for Tanzania that combines medical entomology with advanced data science. Built with FastAPI, React, and PostGIS, it forecasts mosquito populations, optimizes intervention resources, and generates research insights. The system integrates Python/R statistical models and geospatial analytic # VCOM-TZ: AI-Driven Vector Control Optimization Model for Tanzania ## Overview **VCOM-TZ** is an end-to-end decision support system for malaria surveillance and intervention planning in Tanzania. It ingests field surveillance data, predicts mosquito population densities using hybrid forecasting (ARIMA + Prophet), and uses Linear Programming to optimize the allocation of vector control resources (ITNs, IRS, Larvicides). ## Key Features - **Hybrid Forecasting Engine** - Combines Python (Prophet) and R (ARIMA) for high-accuracy density predictions - **Cost-Effectiveness Optimizer** - Linear Programming maximizes cases prevented within budget constraints - **Geospatial Risk Mapping** - Interactive Leaflet choropleth map of Tanzania's 16 regions - **Automated NLP Reporting** - HuggingFace BART model generates surveillance summaries - **Robust Data Pipeline** - PostgreSQL + PostGIS with 2 years of synthetic surveillance data ## Tech Stack | Component | Technologies | |-----------|-------------| | Backend API | FastAPI, Uvicorn, Python 3.11 | | Statistical Modeling | R (forecast, auto.arima), Prophet, SciPy | | Database | PostgreSQL 15 + PostGIS 3.3 | | Frontend | React 18, TypeScript, Leaflet, Chart.js | | DevOps | Docker, Docker Compose | | NLP | HuggingFace Transformers (bart-large-cnn) | ## Architecture ``` VCOM-TZ/ ├── backend/ # FastAPI application │ ├── app/ │ │ ├── api/routes/ # REST endpoints │ │ ├── services/ # Business logic (forecast, optimizer, NLP) │ │ ├── models/ # SQLAlchemy ORM models │ │ ├── schemas/ # Pydantic request/response models │ │ ├── core/ # Configuration, security, dependencies │ │ └── r_scripts/ # R ARIMA integration │ └── tests/ # Pytest suite ├── frontend/ # React TypeScript dashboard │ └── src/ │ ├── components/ # Map, Charts, Optimizer, Reports, Layout │ ├── hooks/ # React Query hooks │ ├── services/ # …