• P-ISSN 0974-6846 E-ISSN 0974-5645

Indian Journal of Science and Technology


Indian Journal of Science and Technology

Year: 2016, Volume: 9, Issue: 10, Pages: 1-5

Original Article

Rule Minimization in Predicting the Preterm Birth Classification using Competitive Co Evolution


Objective: Accurate prediction of pre term birth probability in the deliveries of babies with an effective classifier tool is a big challenging task. This paper talks about a novel competitive co evolution rule prediction classifier for extracting minimum number of rules for identifying the preterm birth. Methods/Analysis: Competitive Co evolution algorithm is applied to the preterm classifier for deriving the pattern governing the classified dataset. In this approach we have used two classes’ namely normal birth dataset and preterm birth dataset as two individuals competing with each other. Fitness of each individual is calculated based on the relative fitness of the other population. Findings: The dataset consist of 1052 records of preterm birth and normal dataset of 1314 each of five attributes. The experimental result shows the total no of rules needed for training and testing is drastically reduced compared to the total rules. The accuracy is also improved to a greater extended by applying the proposed algorithm. Applications/Improvement: The proposed algorithm minimizes the number of training rules to 16 and testing rules to 11 out of total 28 possible rules in the rule set. The accuracy of 0.962938881664499 of correct classification of preterm birth dataset is obtained through method.

Keywords: Co Evolution, Preterm Birth, Reproduction, Rules Extraction


Subscribe now for latest articles and news.