Research & Publications

Sierra Leone's agriculture operates with almost no data-driven decision support, and no published machine learning study has examined the country's crop yields. We ask whether rice yield can be forecast from data Sierra Leone currently has. Using 25 years of FAOSTAT production data (2000-2024) for nine major crops, we train XGBoost, Gradient Boosting, and Random Forest under a strict anti-leakage protocol with expanding-window walk-forward evaluation across seven held-out years, benchmarked against naive persistence. No model trained on crop statistics alone outperforms persistence. Augmenting with free satellite climate data (CHIRPS rainfall, NASA POWER temperature) reverses this result: a climate-only XGBoost reduces forecast error by one third (RMSE 284 vs 428 kg/ha), a gain that holds for a linear model and is robust to excluding the anomalous 2018 season. Early-season (May-June) rainfall is the dominant predictor, implying seasonal yield risk is observable months before harvest. No model anticipated the 2018 collapse, whose origins were institutional rather than climatic. We translate the findings into policy recommendations for Sierra Leone's Feed Salone Strategy, with a fully open-source pipeline.

The United States labor market exhibits a persistent coexistence of high job vacancy rates and prolonged unemployment duration, a pattern that challenges standard labor market theory. This paper argues that a portion of contemporary frictional unemployment is artificially induced by automated recruitment systems that rely on deterministic keyword-based screening. By analyzing system architecture and evaluating semantic matching through experimental simulations, the study quantifies how automated screening design contributes to candidate exclusion and labor market inefficiency.

Automated resume screening systems are now central to large-scale hiring, yet growing evidence suggests that rigid screening logic can exclude qualified candidates before human review. This paper introduces a measurement-based framework for quantifying algorithmic friction in automated recruitment systems. Through controlled simulations comparing keyword-based screening with semantic matching approaches, the study evaluates excess false-negative rejection attributable to semantic mismatch and system design. The results highlight efficiency losses induced by deterministic screening pipelines and provide empirical grounding for assessing hiring system performance.