F011006 超並列データベースTeradataを用いたデータマイニング

抄録

Teradata relational database management system(RDBMS) is providing a data warehouse solution for very large-scale data warehouse from terabytes to tens of petabytes.Teradata RDBMS was born for the parallel processing.The architecture of the Teradata RDBMS makes it possible to divide complicated work into the small work unit and to distribute to the parallelized work unit."Data mining" is called in other word "knowledge discover from database."Data usually accumulated in database.Traditional data mining tools need to extract data from database to their data mining environment.Data volume was increased,data movement cost is very expensive.We had implemented that tool to execute data mining "in-database." "in-database" data-mining usually use variables from transactional data.This is the key point for discovering excellent knowledge.