Performance Analysis on Uncertain Data using Decision Tree

Data uncertainty is common in emerging applications, such as sensor networks, moving object databases, medical and biological fields. Data uncertainty can be caused by various factors including measurements precision limitation. Data uncertainty is inherited in various applications due to different reasons such as outdated sources or imprecise measurement and transmission problems. Classification is one of the most popular data mining techniques. Lot of people used decision tree for data classification and it widely used on certain or precise data. However in this paper we applied on uncertain data which is taken from UCI machine learning repository. This paper proposes a decision tree based classification method on uncertain data. We construct decision tree algorithms by including entropy and information gain, considering the uncertain data intervals. We use some pruning techniques that can improve efficiency of the decision tree and our experiment show that it significantly reduce the tree-construction time.