علي اهنگر

Cancer remains one of the leading causes of illness and mortality worldwide, necessitating effective strategies for early detection and intervention to improve patient outcomes. This study is of high importance as it utilizes machine learning techniques, particularly decision trees, random forests, and multilayer perceptrons, to identify key risk factors and enhance predictive accuracy, ultimately contributing to more personalized and effective healthcare solutions for cancer management. The Golestan cohort was used to eva‎luate the performance of three machine learning models based on occupational and lifestyle factors. Key performance metrics, including accuracy, positive predictive value, sensitivity, and area under the curve (AUC), were employed to assess the effectiveness of the models. The decision tree and random forest models demonstrated higher predictive accuracy, achieving 90.20% and 90.11% accuracy in predicting mortality, respectively, while the MLP model achieved an accuracy of 87.51%. For predicting cancer incidence, both RF and DT models reached near-perfect scores of 99.99%, whereas the MLP recorded a lower accuracy of 97.41% with significant challenges in classification. The findings indicate that RF and DT models are highly effective for predicting cancer and mortality, suggesting their potential application in clinical programs. Continuous optimization and validation of the MLP model are essential to enhance its predictive capabilities and applicability in healthcare settings.

يادگيري ماشين , پيش‌بيني سرطان , نرخ مرگ و مير , سلامت شغلي , سلامت شغلي

Machine Learning , Cancer prediction , Mortality Rate , Occupational Health , predictive analytics

Ali Ahangar

Dr. Mohammad Aliha