Intelligent techniques for finding frequent patterns from large database

Abstract

Frequent Pattern FP mining is a significant and well researched technique of data mining It is used to extract interesting patterns from large database by applying association rules classifier rules correlation rules clustering rules and sequential rules Many researchers have been highly concentrated on FP mining for past years Many efficient pattern mining algorithms have been discovered However these extractions of FPs from large databases are still a challenging and difficult task One of the well known FP mining algorithm is Apriori which address several problems including i incrementally find frequent item sets and associations ii find frequent sub graphs from a set of graphs and iii find subsequences common to several sequences etc Previous apriori techniques have open issues such as high communication cost high response time multiple scanning not acclimate to constantly changing database and too many candidate itemset generation In order to overwhelm these issues this proposed work is designed with two novel algorithms which include FPSSCO Frequent Pattern Sub Spaced Clustering Optimization and SSDPMA Single Scan Distributed Pattern Mining Algorithm These algorithms drastically reduce the existing problems as well as to improve the frequent patterns mining process by establishing links between itemsets. newline