The Eclat algorithm is a typical frequent pattern mining algorithm using vertical data.\nThis study proposes an improved Eclat algorithm called ETPAM, based on the tissue-like P system\nwith active membranes. The active membranes are used to run evolution rules, i.e., object rewriting\nrules, in parallel. Moreover, ETPAM utilizes subsume indices and an early pruning strategy to\nreduce the number of frequent pattern candidates and subsumes. The time complexity of ETPAM is\ndecreased from O(t^2) to O(t) as compared with the original Eclat algorithm through the parallelism\nof the P system. The experimental results using two databases indicate that ETPAM performs very\nwell in mining frequent patterns, and the experimental results using four databases prove that\nETPAM is computationally very efficient as compared with three other existing frequent pattern\nmining algorithms.
Loading....