《Data mining- machine learning》.pdf

Data mining: machine learning, statistics, and databases Heikki Mannila Department of Computer Science University of Helsinki, FIN-00014 Helsinki, Finland E-mail: Heikki.Manni la@cs .helsinki .fi URL: http://www. cs .helsinki .fi/mannila/ Abstract simple example of data mining, namely the problem Know ledge discovery in databases and data min- of discovering asso ciation rules. We present a simple ing aim at semiautomatic to ols for the analysis of algorithm for this task, and show in Section 5 how the large data sets. We give an overview of the area and same ideas can b e used for other typ es of data, to o. present some of the research issues, esp ecially from the Section 6 moves to a generic data mining algorithm, database angle. and discusses some of the architectural issues in data mining systems. Based on these ideas, we consider in 1 Intro duction Section 7 the p ossibilities of sp ecifying KDD tasks in Knowledge discovery in databases (KDD), often a high level language and compiling these sp ecica- called data mining, aims at the discovery of useful tions to ecient discovery algorithms. Section 8 con- information from large collections of data. The dis- siders the p ossibilities of representing large data sets covered