Title :
The Cascade Decision-Tree Improvement Algorithm Based on Unbalanced Data Set
Abstract :
In the past research, the data mining that using single classifier can not obtain satisfactory results. This paper proposed an improved decision-tree classification algorithm M-AdaBoost for solving the customers´ chruning problem. The idea of this algorithm is that using cascaded structure to construct more decision tree classifier based on AdaBoost. This tree have a better classification results according to the experimental results.
Keywords :
customer relationship management; data mining; decision trees; pattern classification; M-AdaBoost; cascade decision-tree improvement algorithm; customer churning problem; data mining; improved decision-tree classification algorithm; unbalanced data set; Algorithm design and analysis; Classification algorithms; Classification tree analysis; Data mining; Decision trees; Error analysis; Logistics; Mobile communication; Mobile computing; Training data; AdaBoost; Data Mining; M-AdaBoost cascade tree;
Conference_Titel :
Communications and Mobile Computing (CMC), 2010 International Conference on
Conference_Location :
Shenzhen
Print_ISBN :
978-1-4244-6327-5
Electronic_ISBN :
978-1-4244-6328-2
DOI :
10.1109/CMC.2010.171
