An Adaptive Method for Mining Frequent Itemsets Based on Apriori And FP Growth Algorithm

Abstract

Mining frequent itemsets is possessed and utilized a wide range of real world applications. Apriori algorithm as well as FP Growth algorithm is one of the most widely used algorithm for mining frequent itemsets. But here is a problem to define minimum support (threshold) to mine frequent itemsets on Apriori and FP Growth algorithm. If minimum support is set to low, too many frequent itemsets will be generated which may cause the Apriori and FP Growth algorithm to become inefficient or even loss of memory. On the other hand, if minimum support is set to too high, less frequent itemsets are found. In this paper, we propose a method to avoid this problem by using Binomial Distribution (BD) to find appropriate minimum support adaptively. It has been helped to mine optimal frequent itemsets so that our proposed method performs better than existing benchmark.