The objective of this Parallel Data Mining Of Association Rules BE Seminar is to generate association rules. This document cover problem of parallel mining of association rules on distributed memory multiprocessor machine. Parallel mining of association rules describe a spectrum of trade-off between memory usage, communication, computation, synchronization and use of problem specific information. “Apriori” parallel algorithm is implemented by which performance is measured and shows a good speed up behavior when compared to sequential implementation.
The growth of scientific, business, and government databases sizes has outpaced our ability to interpret and digest the stored data. This is a need for new generation tools and techniques for automated and intelligent database analysis. These generated tools and techniques are the subjects of data mining.
Data Mining is the process of analyzing data and summarizing into useful information from different perspectives. This Information is used to increase revenue and cut the costs or both. This mining software is used to analyze data, which is one of the best analytical tools.
The primary goal of Data Mining is to obtain knowledge that a user can act on and it is done by building a model of real world from a variety of sources, which include customer’s history, corporate transactions, customer demographic information, and relevant external database.
The major goals of Data Mining are Prediction and Description. Prediction involves usage of fields of the database to predict the unknown values of the variables of interest. Description focuses on finding human-interpretable patterns that describes the data.
The Data Mining approach is generally categorized into six-types. They are Classification, Regression, Time Series, Clustering, Association Rules and Sequence Discovery.
Conclusion:
Implementation of the Apriori Algorithm for Mining of Association rules, it is difficult to carry large amount of data for greater accuracy in a given amount of time. Size up provides sub linear performance as the numbers of transactions are more. It has good speed up performance.
Download Parallel Data Mining Of Association Rules BE / B Tech Seminar.