MODELING INFECTIOUS DISEASE SPREAD USING SEIR COMPARTMENTAL FRAMEWORKS IN LOW-INCOME POPULATIONS

Infectious diseases remain a major challenge for Ghana, reducing productivity, straining budgets, and threatening lives. This study examined how statistical modeling, machine learning, and SEIR extensions shaped outcomes in disease prediction and project resilience between 2020 and 2024. A descriptive design using secondary data from 25 sector-year observations across malaria, cholera, tuberculosis, and COVID-19 guided the analysis. Correlation results showed strong positive links between outcomes and SEIR extensions at 0.81, statistical models at 0.78, and machine learning predictors at 0.74, while contextual constraints had a negative effect at −0.60. Regression confirmed SEIR as the strongest driver with a coefficient of 0.36, followed by statistical models at 0.28 and machine learning at 0.23, with contextual constraints reducing results at −0.20. The model explained 80 percent of variance in project outcomes, validating the framework’s robustness. Findings showed Value-at-Risk thresholds fell from 15 to 11 percent, Monte Carlo worst-case scenarios dropped from 25,000 to 18,500, regression accuracy rose from 70 to 80 percent, ensembles reached 85 percent accuracy, and SEIR reduced R₀ from 2.5 to 1.7. These outcomes imply that quantitative models improve planning, reduce losses, and raise trust in fragile systems, though poor data and weak institutions limit gains. Recommendations urge policymakers to strengthen data and institutional capacity, managers to embed predictive models in dashboards, and educators to train professionals in applied statistical and SEIR modeling.