experimento competencia bosque UPRH

An assignment from Advanced Statistical Learning, taught by Dr. Min Wang at UTSA. The following data sets are analyzed: Chemical manufacturing yield, Fatty acids in food oils, and Customer default records for a credit card company.

An assignment from Advanced Statistical Learning, taught by Dr. Min Wang at UTSA. The following data sets are analyzed: US arrests and Differential gene expression.

This study analyzes the Pima Indians Diabetes Database to develop predictive models for diagnosing diabetes based on diagnostic measurements such as glucose, BMI, insulin levels, and number of pregnancies. The dataset includes medical data from female patients of Pima Indian heritage aged 21 or older. We explored multiple predictive techniques, including logistic regression with and without parameter selection, tree-based methods, k-nearest neighbors, and quadratic discriminant analysis. Models were evaluated using accuracy, ROC-AUC, and other metrics, with cross-validation ensuring robustness. Key predictors like glucose and BMI were consistently significant across models. Random forest and logistic regression emerged as top performers in overall predictive accuracy and interpretability. The study underscores the importance of data-driven approaches in healthcare and highlights actionable insights for improving diabetes diagnostics.