This study is aimed at studying diabetes data using Bayesian Networks and evaluating their performance in detecting Type 1 Diabetes (T1D) and Type 2 Diabetes (T2D). Diabetes dataset was obtained from medical records unit of the Abubakar Tafawa Balewa University Teaching Hospital (ATBUTH) Bauchi, Bauchi State consisting of 569 cases with 8 different variables. Classification of diabetes patients was done by NB and TAN networks; The Networks were trained and tested using 10-fold cross validation and the quality of prediction of these Networks in terms of Sensitivity, Specificity, Area under the ROC (AUR), Kappa Statistic, Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) were evaluated. The results indicate that TAN network proved to be the better network because the method correctly classified 530 (93.15%) and misclassified only 39(6.85%) patients; it also has higher AUR (0.949) and Kappa (0.7869) as well as lower MAE and RMSE of 0.1032 and 0.2384 respectively.
Key words: Nodes, Directed Acyclic Graph, Edge, Markov blanket, ROC Curve
|
