Efficient Algorithms for Mining Closed Frequent Itemset and Generating Rare Association Rules fromUncertain Databases
Journal Title: International Journal of Scientific Research and Management - Year 2013, Vol 1, Issue 2
Abstract
Dealing with uncertainty has gained increasing attention in the past few years in both static and streaming data management and mining. Data uncertainty is inherent in emerging applications such as location - based services, sensor monitoring syst ems and data integration. Specifically some privacy – preserving data stream mining applications like banking sectors operate on uncertain data and they need to mine frequent itemsets. There are many existing works on mining frequent itemsets from uncertain databases. They use the expected support of an itemset from an uncertain database to define whether it is frequent or not. Itemsets are considered frequent if its expected support exceeds the minimum support threshold. There are two main problems with this approach of mining frequent itemsets from uncertain databases. The first problem is that often too many frequent itemsets are generated many of which may prove to be irrelevant. The second is that only expected support is calculated to define frequent ite msets as being those whose expected support exceeds the minimum support. To overcome both these problems closed frequent itemset mining can be used. An itemset X is said to be closed, if and only if none of its superset has the same support. The mining pro cess can be done by using Poisson Binomial Distribution. Closed frequent itemsets are extracted by an approximate mining algorithm which can effectively and accurately discover closed frequent itemsets in large uncertain databases. It can be extended to ge nerate the rare association rules within the given support and confidence framework. In a practical scenario, there can be some itemsets which have significance and can provide useful knowledge but their support counts are relatively less. To extract the r elationship among those itemsets having less frequency of occurrences, rare association rule mining came into existence.
Authors and Affiliations
A. L. Greenie Geevlin
Keywords
Related Articles
- EP ID EP206419
- DOI -
- Views 136
- Downloads 0
C OGNITIVE RADIO
The main aim is to exploit sparsity of the spectrum. This provides a more efficient use of the spectrum as well as reduce power consumption thereby enabling high priority communications to take precedence i...
Neurofinance: An interdisciplinary Science of correlating the Neurology and the Individual Investment Patterns of Buyers
Present article is a conceptual research to document the importance Neurofinance . Researchers have spent a couple of days to understand the concept and approach for the Investment Patterns of Buyers in any sector . This...
Effectiveness o f Relax ation Handheld Fingertechnique and Benson Relaxation t o the Changes Level of Post Operative Pain Sectio Caesarea in Rsi Sakinah Mojokerto
Pain after major or minor surgery will impact the anatomical and physiological changes in patients. Giving anesthetic is not always sufficient to eliminate the s ensation of pain because the pain response of each patient...
Communication A Tool To Build Relationship Quality: Evidence From Indian Retail Banks
The purpose of this paper is to study the effect of communication on relationship quality dimensions such as trust and commitment in retail banks. Data were gathered from the retail bank customers in Che...
Variation of Refractive Index of PMMA with temperature an d different doping % of Tio 2
A simple and reliable method of measuring the refractive index of thin film is reported in the present paper. In this work, an experimental approach of Digital Abbe Spectrometer was used for measuring refrac...