Volume & Issue no: Volume 5, Issue 6, November - December 2016
____________________________________________________________________________________________________
Title: |
Fast Computation of Probabilistic Frequent Patterns in Uncertain Databases |
Author Name: |
Dr. Telkapalli Murali Krishna , Amsalu Tomas |
Abstract: |
Abstract
In practice the data in the database is always stochastic by
nature. The probabilistic assumption of data in deterministic
databases can be also formulated as uncertainty of data with a
value 1.0. The uncertain nature of the data may come from
dynamic error-rapid change in the environment or scene of
the data source, drifting in reading value like temperature,
and noise of the equipment, distance etc. This nature can be
seen in both object and attribute level uncertainties. In
general, preprocessing in data warehouse− typically ETL, data
integration, data granularity, ambiguous entities or missing
data values can be main causes of the uncertain database.
Mining such databases needs more attention in computing
support count of itemset that are uncertain with their
corresponding existential probabilities. This paper focuses on
faster and efficient method to compute probabilistic support
count of an itemset using divide and conquer approach as
transforms on a sample dataset. The sampling technique has
been seen as a crucial technique to minimize the complexity
as the size of dataset and support computation are
proportional. The probabilistic support can be found from
inverse transform of the interpolated values of probabilistic
support function. The performance of this approach is
compared with the existing probabilistic pattern mining
algorithms.
Keywords: uncertainty, computing support,probabilistic
pattern mining, transforms, |
Cite this article: |
Dr. Telkapalli Murali Krishna , Amsalu Tomas , "
Fast Computation of Probabilistic Frequent Patterns in Uncertain Databases" , International Journal of Emerging Trends & Technology in Computer Science (IJETTCS) ,
Volume 5, Issue 6, November - December 2016 , pp.
061-066 , ISSN 2278-6856.
|
| Full Text [PDF] Back to Current Issue | |
NOTE: Authors note that paper cannot be withdrawn at any condition once it is accepted. The Team of IJETTCS advise you, do not submit same article to the multiple journals simultaneously. This may create a problem for you. Please wait for review report which will take maximum 01 to 02 week.