Department of Mathematics, Statistics and Actuarial Science, Faculty of Applied Science, Takoradi Technical University, P. O. Box 256, Takoradi, Western Region, Ghana
Auto-generated profile, based on publicly available information
Emmanuel Ayitey
Department of Mathematics, Statistics and Actuarial Science, Faculty of Applied Science, Takoradi Technical University, P. O. Box 256, Takoradi, Western Region, Ghana
This methodological study uses a controlled simulation, not a new empirical assessment of Pra River Basin water quality, to examine two validation pitfalls that inflate apparent machine-learning performance in predicting a Water Quality Index (WQI): target circularity, where a model is scored on the variables used to construct the index, and optimistic within-domain validation, where random splits mask poor performance outside the sampled domain. A synthetic, 150-record benchmark was generated for the Birim, Offin and Pra sub-basins, with WQI calibrated as a near-deterministic function of pH, TSS, TDS and electrical conductivity. A full-feature model was compared with an auxiliary-only model using XGBoost, Random Forest, linear and Ridge regression, assessed via five-fold and leave-one-sub-basin-out validation. Full-feature XGBoost performed well within sub-basins (R² = 0.729–0.830) but dropped to negative R² when index-defining variables were removed. Transfer was weak for every model: XGBoost produced R² of 0.078 to 22.982 − − out-of-basin, and Ridge showed comparably negative transfer, confirming the failure was not confined to tree ensembles. Reduced auxiliary models did not recover transferable skill. Findings show strong local WQI prediction can coexist with poor geographic transfer and is not resolved by using linear rather than tree-based models, supporting validation aligned with intended deployment.
Background This study explores the dynamics of prostate cancer progression, aiming to understand how different stages of the disease interact over time. By constructing two-state, three-state, and four-state compartmental models, the research seeks to provide insights into the impact of disease transition rates and mortality on the spread of prostate cancer. Each model represents a stepwise progression of cancer, highlighting how early detection and treatment could influence disease management and patient outcomes. Methods The study employs compartmental modeling to simulate transitions between healthy individuals and those in various stages of prostate cancer in Ghana. The models range from a two-state system, which captures the basic transition from healthy to cancerous states, to a four-state model that includes early-stage, advanced-stage, and treatment compartments. Parameter estimation techniques and sensitivity analyses were used to assess the impact of transition and mortality rates. Statistical methods, including sensitivity and parameter estimation analyses, were applied to understand how variations in these rates affect overall cancer dynamics. Results The two-state model showed a clear inverse relationship between healthy individuals and the incidence of prostate cancer, suggesting a rapid decline in healthy individuals as cancer cases increase. The three-state model provided a more nuanced understanding by differentiating between early and advanced stages, demonstrating the importance of early detection. The four-state model incorporated the effects of treatment, revealing that effective intervention could significantly alter the progression dynamics. Sensitivity analyses highlighted the influence of early-stage detection and the treatment stage on controlling the spread of prostate cancer. Conclusions This research contributes to the importance of early detection and timely treatment in managing prostate cancer. The findings suggest that intervention strategies should be tailored based on disease stage. By refining our understanding of prostate cancer progression through compartmental modeling, this study offers a mathematical model for further research and provides potential pathways for optimizing treatment approaches to enhance patient outcomes.