Algorithms for Mining Frequent Patterns A Comparative Study

Abstract: Mining frequent patterns are one of the most important research topics in data mining. The function is to mine the transactional data which describes the behaviour of the transaction. In an online business or in an online shopping the customers can purchase items together. Frequent patterns are patterns such as item sets, subsequence or substructures that appear in a data set frequently. Many efficient algorithms were developed based on the data structure and the processing scheme. The mining of most efficient algorithms such as Apriori and FP Growth were implemented here. In this paper we propose the efficient algorithms (Apriori and FP Growth) used to mine the frequent patterns. The Apriori algorithm generates candidate set during each pass. It reduces the dataset by discarding the infrequent itemsets that do not meet the minimum threshold from the candidate sets. To avoid the generation of candidate set which is expensive the FP Growth algorithm is used to mine the database.

1