Decision Tree Approach for Classifying Uncertain Data

Decision tree is powerful and popular tool for classification and prediction in uncertainty data. This study proposes a decision tree based classification system for uncertain data. The uncertain data means lack of certainty. Data uncertainty comes by different parameters including sensor error, network latency measurements precision limitation and multiple repeated measurements.It is found that decision tree classifier gives more accurate result if take “complete information” of data set is taken. In this paper, the traditional decision tree algorithm is modified by combining firefly and weighted entropy measure. Results obtained from three UCI repositories demonstrate that t he proposed measure results in decision trees that are more compact with classification accuracy that is comparable to that obtained using popular nose splitting measure. The simulation result demonstrates that the proposed system gives better results for uncertain data and it is computationally efficient in terms of accuracy.