TwoFold Frisky Algorithm (TFFA): A Fast Frequent Itemset Algorithm

摘要

Finding frequent itemsets and generation of association rules by using frequent items plays an important role in the field of data mining. Many algorithms were proposed to get frequent itemsets, but the most popular algorithm is Apriori which is implemented on the horizontal database. In which the method frequently scans the database and returns the flood of candidates which are the significant disadvantages. A novel algorithm based on the vertical database was introduced, which overcomes the disadvantages of Apriori. The proposed algorithm discards the calculation of a few frequent itemsets by taking the next maximum itemset in the process of generating maximal frequent itemset. That means when the length of a frequent itemset with the higher value in powers of two appears then it neglects the lower valued itemsets though it is in powers of two. The simulation results were compared with Apriori and FP-Growth algorithms. It was shown that the novel implementation performed better than Apriori and FP-Growth.