ID3 Decision Tree Classification: An Algorithmic Perspective based on Error rate

Abstract

Classification in Data mining is a very important approach that is widely used in all the applications including medical diagnoses, agriculture, and other decision making systems. Data mining deals primarily with classification due to dynamic varieties of datasets available online today. Decision tree based classification is the foundation of all the classification algorithms and is extensively used by experts in all types of research. As ID3 decision tree algorithm has been used popularly for classification, this proposed work focuses on the implementation of ID3 algorithm with different standard UCI datasets and are also analyzed using statistical measures. The Error rate determines the misclassification of an algorithm and the splitting attribute. Hence, the ID3 algorithmic perspective was carried out in this proposed work by analyzing the error rate.