01669naa a2200241 a 450000100080000000500110000800800410001910000200006024500430008026000090012352010200013265000140115265000180116665300150118465300160119965300310121565300240124665300230127065300260129370000180131970000170133777300730135415555482001-10-18       2000    bl ---       0--  u    #d1 aPRADO, H. A. do  aClustering algorithms for data mining.  c2000  aKnowledge Discovery from Databases (KDD) can be seen as a set of computer-aided Knowledge discovery techniques. Research in this field has heen carried out according to two main dimension: simplifying and scaling of the whole process to cope with very large databases. The first efforts of KDD research community addressed strongly the task of prediction, e.g., regression and classification. The aim of this task is to fit a model over a data set for which it is Known some interesting feature in order to predict the same feature for a new case. Recently, researchers have focussed the task of description  that includes clustering and finding associative rules. In using the descriptive approach one is interested in figures out how a set of objects are organized in the space of their dimensions. This paper presents a study on clustering methods focussing, in particular, those that have been the target of research efforts in the KDD realm. Our objective is to state a well-founded basis for forthcoming work.  adatabases  aBase de Dados  aClustering  aData mining  aDescoberta de conhecimento  aKnowledge discovery  aMineracao de dados  aUnsupervised learning1 aHIRTLE, S. C.1 aENGEL, P. M.  tRevista Tecnologia da Informacao, Brasiliagv.2, n.1, p.51-58, 2000.