文件名称：Decision Trees for Business Intelligence and Data Mining
文件大小：4.08MB
文件格式：PDF
更新时间：2013-03-14 02:24:15
decision tree sas
Data has an important and unique role to play in modern civilization: in addition to its historic role as the raw material of the scientific method, it has gained increasing recognition as a key ingredient of modern industrial and business engineering. Our reliance on data—and the role that it can play in the discovery and confirmation of science, engineering, business, and social knowledge in a range of areas—is central to our view of the world as we know it. Many techniques have evolved to consume data as raw material in the service of producing information and knowledge, often to confirm our hunches about how things work and to create new ways of doing things. Recently, many of these discovery techniques have been assembled into the general approaches of business intelligence and data mining. Business intelligence provides a process and a framework to place data display and data analysis capabilities in the hands of frontline business users and business analysts. Data mining is a more specialized field of practice that uses a variety of computer-mediated tools and techniques to extract trends, patterns, and relationships from data. These trends, patterns, and relationships are often more subtle or complex than the relationships that are normally presented in a business intelligence context. Consequently, business intelligence and data mining are highly complementary approaches to exposing the full range of information and knowledge that is contained in data. Some data mining techniques trace their roots to the origins of the scientific method and such statistical techniques as hypothesis testing and linear regression. Other techniques, such as neural networks, emerged out of relatively recent investigations in cognitive science: how does the human brain work? Can we reengineer its principles of operation as a software program? Other techniques, such as cluster analysis, evolved out of a range of disciplines rooted in the frameworks of scientific discovery and engineering power and practicality.