Nicolekr7/Agent-Based-Model-Streamlit

A climate-linked Agent-Based Modeling system for predicting malaria outbreaks in Kisumu County, Kenya. Includes data preparation, SEIR-based ABM simulation engine, and a fully interactive Streamlit dashboard for visualization and outbreak forecasting. **Malaria Outbreak Prediction Using Agent-Based Modeling (ABM)** An Interactive Climate-Linked Simulation System for Kisumu County, Kenya This repository contains the full implementation of an Agent-Based Modeling (ABM) system for simulating and predicting malaria outbreaks in Kisumu County, Kenya. It includes data preprocessing, outbreak simulation, climate integration, visualization, and deployment through a fully interactive Streamlit dashboard. **Project Overview** Malaria remains a significant public health challenge in the Lake Victoria region. This project applies a climate-sensitive SEIR-based ABM to simulate malaria transmission dynamics at the micro-level, allowing users to explore how changes in rainfall, temperature, and human dynamics influence outbreak patterns. The model uses: SEIR agent transitions (Susceptible → Exposed → Infected → Recovered) Climate-driven infection probability Individual-level variation in exposure, recovery, and immunity Visualization of emergent epidemic peaks **Features** **Fully Interactive Streamlit Dashboard** Upload malaria–climate dataset Visualize rainfall, temperature, malaria trends Run baseline ABM or climate-linked ABM Adjust model parameters in real-time Watch animated epidemic curves View 2D agent grid showing S/E/I/R state distribution Generate and download outbreak simulation results **Climate-Driven Infection Modeling** Rainfall and temperature weighted transmission Seasonal cyclic behavior Effects of climate anomalies on outbreak timing **Clean and Modern UI** Fully tabbed layout Real-time updates Streamlit Cloud deployment support **Dataset** The model uses a synthetic weekly dataset (2015–2022) approximating: Rainfall (mm) Temperature (°C) Malaria cases Normalized rainfall & temperature The dataset can be uploaded via the Streamlit UI or included in the repository.