Title :
The rule-matching algorithm of decision tree attribute reduction
Author :
Li, Yan ; Li, Fa-chao ; Li, Yun-hong
Author_Institution :
Sch. of Sci., Hebei Univ. of Sci. & Technol., Shijiazhuang
Abstract :
The attribute reduction of information system can improve the accuracy of knowledge discovery, machine learning, etc. and it also can improve the efficiency. This paper proposes an attribute testing reduction algorithm, the algorithm can make the information system retain as few as attributes under the condition that maintains the original style, it can not only save much time for the later system handling, but reduce time complexity of algorithm and improve the computation precision.
Keywords :
data mining; decision trees; learning (artificial intelligence); rough set theory; computation precision; decision tree attribute reduction; information system; knowledge discovery; machine learning; rule-matching algorithm; time complexity; Classification tree analysis; Decision trees; Machine learning; Machine learning algorithms; Management information systems; Partitioning algorithms; Pattern analysis; Pattern recognition; Set theory; Wavelet analysis; Attribute reduction; Decision tree; Rough set; Rule matching degree; Threshold;
Conference_Titel :
Wavelet Analysis and Pattern Recognition, 2008. ICWAPR '08. International Conference on
Conference_Location :
Hong Kong
Print_ISBN :
978-1-4244-2238-8
Electronic_ISBN :
978-1-4244-2239-5
DOI :
10.1109/ICWAPR.2008.4635898
