Applying Apriori Algorithm with Parallel Computing Concepts in Forest Fire Prediction - A Case study

  • GOPICHAND G , SENTHILNATHAN PALANIAPPAN*, NARESH K, ANITHA K , RAMANI SELVANAMBI

Abstract

Apriori algorithm is a popular association rule mining algorithm which helps in finding various frequent item sets in the database. The constraints for finding these item sets are given by the user in terms of support - calculated by the ratio of transactions in which an item set appears, and confidence –estimated by the ratio of transactions with an item set, which also consists of another item set. The problem with this algorithm is highly iterative and efficiency rapidly decreases with size increase or dimension of the dataset. In proposed model increases its efficiency with the help of OpenMP threads. It use data decomposition to split the transaction database into various parts, each taken by a thread to find the support count of all the candidate item sets for all the transactions assigned to that particular thread. for example of the application, proposed models used to determine the probability of the occurrence of a forest fire. Here, the transaction database can consist of various occurrences of natural phenomena, in which a few transactions also have the forest fire phenomenon, which means that it has occurred in the presence of the other itemsets in the transaction. Hence, if a new transaction is taken from the user, then the probability (or confidence) that a forest fire occurs, given this transaction, is calculated. The output for a 12 iteration based Apriori Algorithm in Serial implementation is 9.766 seconds whereas for parallel implementation is 6.86 seconds. This increases the efficiency of algorithm by 42%.

 

Published
2020-06-01
How to Cite
GOPICHAND G , SENTHILNATHAN PALANIAPPAN*, NARESH K, ANITHA K , RAMANI SELVANAMBI. (2020). Applying Apriori Algorithm with Parallel Computing Concepts in Forest Fire Prediction - A Case study . International Journal of Advanced Science and Technology, 29(7), 4873-1114. Retrieved from http://sersc.org/journals/index.php/IJAST/article/view/23532
Section
Articles