Apriori Algorithm against Fp Growth Algorithm: A Comparative Study of Data Mining Algorithms

Abstract

Finding frequent patterns from vast databases is a difficult issue in Data Mining, and numerous studies are conducted on a regular basis. In this paper, a comparison is conducted between the Apriori algorithm, which uses candidate set generation and testing, and the FP growth method, which does not employ candidate set formation. When the apriori algorithm discovers a frequent item set, all of its subsets must likewise be frequent. The apriori algorithm generates candidate item sets and determines how common they are. Pattern fragment growth is used in the FP growth technique to mine frequent patterns from huge databases. For storing critical and compressed information on frequent patterns, an extended prefix tree structure is used. FP growth finds frequent item sets without generating candidate item sets. In data mining, association rule mining is a well-known and well-researched technique for discovering surprising correlations between variables in huge databases. Various Data Mining techniques, such as grouping, clustering, and prediction, are referred to as the rule of association. The purpose of this work is to compare the capabilities of the Apriori and Frequent Pattern (FP) growth algorithms. The FP-growth method outperforms the Apriori algorithm.

Keywords: Apriori, FP-growth, Support, Confidence Apriori Algorithm, FP-Growth Algorithm, FP-Tree Structure