Hybrid classifier for increasing accuracy of fitness data set
Abhishek Lal, Suthikshn Kumar · 2017
Machine learning and Data mining techniques are rapidly establishing themselves in medical and health care fields. This paper addresses a similar issue where the fitness of an individual can be predicted by analyzing few attributes associated with that individual. A hybrid classifier algorithm is developed by merging Decision Tree and Naïve Bayes algorithms which will classify the Fitness data set. The nature of the data set is such that the classification accuracy does not reach the desired levels with individual classifiers but when classification is carried out with the hybrid classifier the classification accuracy is enhanced by 15.79% as compared to Decision Tree and 3.6% when compared with Naïve Bayes classifier. This concept displays an approach where tailor made hybrid data mining algorithms can be developed for classifying particular data sets which have unique nature and require specific classification requirements for better accuracy.