Evolving a Bayesian classifier for ECG-based age classification in medical applications

Abstract. Objective: Age classification of patients based on information extracted from electrocardiograms (ECG's). The scope of this work is to develop and compare the performance of Bayesian classifiers. Methods and Material: We present a medical patient classification methodology using a genetically evolved Bayesian network classifier and biological signal characteristics. Patient age classification is performed based on statistical features extracted from electrocardiogram signals. The continuous signal feature variables are converted to a discrete symbolic form based on thresholding to lower the dimensionality of the signal allowing for smaller conditional probability tables to be calculated for the classifier. Two methods of network discovery from data were developed and compared: the first using a greedy hill-climb search and the second method based on evolutionary computing using a genetic algorithm (GA). Results and Conclusions: Performance of both Bayesian network discov...

• Type: Journal
• Year: 2008
• Where: ASC
• Authors: M. Wiggins, A. Saad, Brian Litt, George J. Vachtsevanos
• Keywords: ASC 2006, ASC 2008, Bayesian Classifiers, Evolved Bayesian Network, Naive Bayesian Classifiers