Data Mining of Access to Tetanus Toxoid Immunization Among Women of Childbearing Age in Ethiopia

• DOI: 10.11648/j.mlr.20170202.12
• Publication Date: 2017-03-09
• Authors: Kedir Hussein Abegaz; Emiru Merdassa At...
• Keywords: Neonatal tetanus; Toxoid

Tetanus toxoid (TT) vaccine is given to women of childbearing age prevent neonatal tetanus and maternal mortality attributed tetanus. Globally, responsible for 5% deaths 14% annually. Data mining the process discovering interesting patterns knowledge from large amounts data. Thus, aim this study was identify best classifier, predict pattern TT data set using algorithms technique. The were Toxoid Ethiopian Demographic Health Survey (EDHS) 2011, analyzed Knowledge discovery Selection, Processing, Transforming, mining, interpretation. WEKA 3.6.1 tool used classification, clustering, association attribute selection. accuracy rate classifiers on training relatively higher than test multilayer perceptron classifier in our toxoid. In cross-validation with 10 folds, correctly classified are by naive Bayesian 63.30% least accurate k-nearest neighbor 60.52%. Single instance Naive done creating 1, 2, 3, 4 instance, three them predicted but one incorrectly classified. maximum confidence attained general 0.98. But, class attribute, it 0.72. literacy status mother has high information gain value 0.046. As a conclusion, algorithm based vaccination an 67.28% total time taken build model at 0.01 seconds. Multilayer lowest average error 32.72% compared others. These results suggest that among machine learning tested, potential significantly improve conventional classification methods use EDHS