Browsing by Author "Naziwa, Aisha"

Filter results by typing the first few letters
Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Comparison of the Naive Bayes Classifier and Instance Based Learner in Classifying Upper Gastrointestinal Bleeding
    (MATEMATIKA, 2013) Naziwa, Aisha; Mohd, Bakri Adam; Shamarina, Shohaimi
    Upper gastrointestinal bleeding is a medical emergence that results in high medical costs and death. Management of this disease requires ascertaining the cause of bleeding. The cause of bleeding is classified into esophageal and gastric causes. Based on health survey data, this study compares the accuracy of the naive Bayes classifier and an instance based learner in the classification of the cause of bleeding. The two classifiers are learned and trained on data collected from patients admitted for upper gastrointestinal bleeding. The naive Bayes classifier achieves a classification accuracy of 71% accuracy compared to 68% of the instance based learner
  • Thumbnail Image
    Item
    Effect of Missing Value Methods on Bayesian Network Classification of Hepatitis Data
    (International Journal of Computer Science and Telecommunication, 2013-06) Naziwa, Aisha; Mohd, Bakri Adam; Shamarina, Shohaimi
    Missing value imputation methods are widely used in solving missing value problems during statistical analysis. For classification tasks, these imputation methods can affect the accuracy of the Bayesian network classifiers. This paper study’s the effect of missing value treatment on the prediction accuracy of four Bayesian network classifiers used to predict death in acute chronic Hepatitis patients. Missing data was imputed using nine methods which include, replacing with most common attribute, support vector machine imputation (SVMI), K-nearest neighbor (KNNI), Fuzzy K-means Clustering (FKMI), K-means Clustering Imputation (KMI), Weighted imputation with K-Nearest Neighbor (WKNNI), regularized expectation maximization (EM), singular value decomposition (SVDI), and local least squares imputation (LLSI). The classification accuracy of the naive Bayes (NB), tree augmented naive Bayes (TAN), boosted augmented naive Bayes (BAN) and general Bayes network classifiers (GBN) were recorded. The SVMI and LLSI methods improved the classification accuracy of the classifiers. The method of ignoring missing values was better than seven of the imputation methods. Among the classifiers, the TAN achieved the best average classification accuracy of 86.3% followed by BAN with 85.1%.