Feature Selection as an Improving Step for Decision Tree Construction
Mahdi Esmaeili, Fazekas Gabor · 2011
The removal of irrelevant or redundant attributes could benefit us in making decisions and analyzing data efficiently. Feature Selection is one of the most important and frequently used techniques in data preprocessing for data mining. In this paper, special attention is made on feature selection for classification with labeled data. Here an algorithm is used that arranges attributes based on their importance using two independent criteria. Then, the arranged attributes can be used as input one simple and powerful algorithm for construction decision tree (Oblivious Tree). Results indicate that this decision tree using featured selected by proposed algorithm outperformed decision tree without feature selection. From the experimental results, it is observed that, this method generates smaller tree having an acceptable accuracy.