Descriptors :
Data mining , frequent pattern , expected support , tree structure , existential probability , uncertain data
Abstract :
Most of the data mining algorithms were designed to mine the frequent pattern from precise data. However, uncertainty exists in many real life situations such as sensor network and privacy preserving applications. To extract meaningful information from uncertain data a number of frequent pattern mining algorithms have been proposed. While dealing with uncertain data U-Apriori, UF-growth, UFP-growth, UH-mine, PUF-growth, TPC-growth algorithm are examples of existing frequent pattern mining algorithms, which utilize different approaches to mine frequent pattern. One important observation is that algorithms behave completely different in the uncertain database as compared to the precise database due of the inclusion of probability value. In this survey paper, a number of algorithms have been analyzed for finding the frequent pattern from uncertain database. The analysis is represented in the form of comparative study of following algorithm: U-Apriori, UF-growth, UFP-growth, UH-mine, and PUF-growth, TPC-growth algorithm on the basis on various parameters such as database scan, running time, memory utilization and storage structure. The survey paper also focuses on the advantage and limitation of each algorithm.