machine learning and data mining machine learning algorithms文档精选.pdfVIP

G e n e r a l D o m a i n s Machine Learning and Data Mining Machine learning algorithms enable discovery of important “regularities” in large data sets. Over the past Tom M. Mitchell decade, many organizations have begun to routinely cap- ture huge volumes of historical data describing their operations, products, and customers. At the same time, scientists and L engineers in many fields have been captur- M T H . X ing increasingly complex experimental data E D N I / M O sets, such as gigabytes of functional mag- C . E C A P S Y netic resonance imaging (MRI) data S A E . E R U describing brain activity in humans. The T U F . V O H field of data mining addresses the question T R O N P / R of how best to use this historical data to E V O H T R discover general regularities and improve O N P I L I H P the process of making decisions. COMMUNICATIONS OF THE ACM November 1999/Vol. 42, No. 11 31 he increasing interest in Figure 1. Data mining application. A historical set of 9,714 medical data mining, or the use of records describes pregnant women over time. The top portion is a historical data to discover typical patient record (“?” indicates the feature value is unknown). Tregularities and improve The task for the algorithm is to discover rules that predict which future decisions, follows from the future patients will be at high risk of requiring an emergency conf