A Parallel Algorithm for Partial Multiple Periodic Patterns Mining with Minimum Inter-Processor Communication

• Main Authors: Jia-Ming Lee, 李家銘
• Other Authors: Guanling Lee
• Format: Others
• Language: en_US
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/70119644272000704783

碩士 === 國立東華大學 === 資訊工程學系 === 92 === Partial periodic patterns mining is a very interesting topic in data mining problem. It is widely used in the market analysis, such as stock management and sale management, etc. However, as the amount of data increases, the scalability of data mining algorithm has become a very important objective. To improve the scalability, in recent years, the concept of parallel computing has been applied on general data mining algorithm. In this thesis, the problem of mining partial multiple periodic patterns under the parallel computing environment is discussed. To reduce the communication cost between the processors, in our approach, the independent property of prime number is employed to classify partial periodic patterns into several independent sets. Moreover, a novel method of distributing mining tasks among the processors is proposed. A set of simulations is also performed to show the benefit of our approach.