Classify Uncertain Data with Decision Tree

This demo presents a decision tree based classificationsystem for uncertain data. Decision tree is a commonlyused data classification technique. Tree learning algorithms cangenerate decision tree models from a training data set. Whenworking on uncertain data or probabilistic data, the learning andprediction algorithms need handle the uncertainty cautiously, orelse the decision tree could be unreliable and prediction resultsmay be wrong. In this demo,we will present DTU, a new decisiontree based classification and prediction system for uncertaindata. This tool uses new measures for constructing, pruningand optimizing decision tree. These new measures are computedconsidering uncertain data probability distribution functions.Based on the new measures, the optimal splitting attributes andsplitting values can be identified and used in the decision tree.We will show in this demo that DTU can process various typesof uncertainties and it has satisfactory classification performanceeven when data is highly uncertain.

Authors and Affiliations

1. Indiana University Purdue University, Indianapolis, USA

   Biao Qin, Yuni Xia, Rakesh Sathyesh & Jiaqi Ge

2. Purdue University, West Lafayette, USA

   Sunil Probhakar

Editors and Affiliations

1. Department of Systems Engineering and Engineering Management, The Chinese University of Hong Kong, Shatin, N.T., Hong Kong, China

   Jeffrey Xu Yu

2. Department of Computer Science, Korea Advanced Institute of Science and Technology (KAIST), 291 Daehak-ro (373-1 Guseong-don), 305-701, Yuseong-gu, Daejeon, Korea

   Myoung Ho Kim

3. Institute for Computer Science and Business Information Systems (ICB), University of Duisburg-Essen, Schützenbahn 70, 45117, Essen, Germany

   Rainer Unland

Reprints and permissions

© 2011 Springer-Verlag Berlin Heidelberg

Policies and ethics