Title: Dependence of the data mining techniques on the decision problem nature
Authors: Hoda A. Abdelhafez; Kamal Abdel-Raouf ElDahshan
Faculty of Computers & Informatics, Information Systems & Decision Support Department, Suez Canal University, Old Campus, ElShikh Zayed, Ismaila, Egypt
Faculty of Sciences, Computer Science and Information Systems Department, AlAzhar University, Nasr City, Cairo, Egypt
Abstract: Data mining is a valuable tool for many industries, including finance, manufacturing, retail, medical health, insurance and telecommunications. The aim of this paper is to investigate and demonstrate that each class of decision problems needs a suitable or set of data mining techniques. Each category of decision problems applies some data mining techniques that may differ from the other categories. These techniques are determined based on the type of the problem, the scale of the dataset, as well as the dynamic changes of these data. The results illustrated that the choice of the data mining technique is decision problem-dependent meaning that for each class of decision problems there is a suitable class of data mining techniques.
Keywords: data mining; decision problems; problem classification; data science; fraud detection; customer behaviour; risk management; fault patterns; customer churn; context.
Int. J. of Data Science, 2015 Vol.1, No.2, pp.103 - 117
Available online: 13 Oct 2015