FP-Growth Tree Based Algorithms Analysis: CP-Tree and K Map

Abstract

We propose a novel frequent-pattern tree (FP-tree) structure; our performance study shows that the FP-growth method is efficient and scalable for mining both long and short frequent patterns, and is about an order of magnitude faster than the Apriori algorithm and also faster than some recently reported new frequent-pattern mining methods. FP-tree method is efficient algorithm in association mining to mine frequent patterns in data mining, in spite of long or short frequent data patterns. By using compact best tree structure and partitioning-based and divide-and-conquer data mining searching method, it can be reduces the costs searchsubstantially .it just as the analysis multi-CPU or reduce computer memory to solve problem. But this approach can be apparently decrease the costs for exchanging and combining control information and the algorithm complexity is also greatly decreased, solve this problem efficiently. Even if main adopting multi-CPU technique, raising the requirement is basically hardware, best performanceimprovement is still to be limited. Is there any other way that most one may it can reduce these costs in FP-tree construction, performance best improvement is still limited.