Enhancing the Prediction Accuracy for Cardiotocography (CTG) using Firefly Algorithm and Naive Bayesian Classifier
Noora Jamal Ali Kadhim, Jameel Kadhim Abed
Abstract
Abstract Recently, there is a huge amount of data accessible in the field of medicine that enables physicians diagnose diseases when analyzed. Data mining technology can be used to obtain knowledge from these medical data in order to make disease prediction accurate and easier. In this study, cardiotocography (CTG) data is analyzed using an integrated Naive Bayesian classifier nbc with firefly algorithm. Firefly algorithm is suggested to find the most relevant subset of features, which maximize the performance accuracy of nbc and minimize the time required for classification process. It was discovered that the nbc was capable of defining the Normal, Suspicious and Pathological state of the type of the CTG data with very good classification accuracy. the proposed method achieved accuracy with (86.547%).