Sunday-Oladokun/Advanced-Geospatial-Analysis-for-Election-Integrity-in-Oyo-State-Nigeria.

# Advanced-Geospatial-Analysis-for-Election-Integrity-in-Oyo-State-Nigeria. This project utilizes geospatial analysis, statistical modeling, and machine learning to examine the 2023 Presidential Election results in Oyo State, Nigeria, and identify potential electoral irregularities. ## Key Objectives * Detect statistical outliers in voting results using spatial autocorrelation measures. * Identify geographic clusters of irregularities through DBSCAN clustering. * Validate anomalies using machine learning (Isolation Forest). * Contextualize findings through historical and demographic comparisons. ## Methodology The analysis involves a four-stage pipeline: 1. **Data Preparation:** Geocoding polling units using Google Maps API and cleaning the dataset. 2. **Spatial Analysis:** Applying HDBSCAN clustering to group polling units and using Local Moran's I and Getis-Ord Gi\* for outlier detection. 3. **Machine Learning Validation:** Training an Isolation Forest model to detect global anomalies. 4. **Contextualization:** Comparing 2023 results with historical data and mapping anomalies to socio-economic indicators. ## Results Key deliverables include: * Geocoded dataset with outlier scores. * Prioritized list of high-risk polling units. * Interactive dashboard for visualizing clusters and anomalies : (app.powerbi.com) * Analytical report with findings and recommendations. ## Impact This project provides a framework for electoral authorities to efficiently audit irregularities and enhance transparency in the electoral process.