Hybrid Fuzzy Frequency Monetary Algorithm (Hffma) And Fuzzy Particle Swarm Optimization And Ranking Algorithm Using Combination Of Sequential Pattern Mining

Abstract

An efficient, Hybrid Fuzzy Frequency Monetary Algorithm (HFFMA) algorithm is proposed in this Chapter to handle the maintenance problem of HFFMA-sequential patterns. An updated HFFMA-tree is built using the HFFMA- sequential patterns obtained from the static database to control the dynamic nature of data updating process and deletion process into the sequential pattern mining problem. Subsequently, the database gets updated from the distributed database that may be static, inserted, or deleted. .  There are two user-specified reference values required by our proposed approach. One indicates the frequency of expected patterns, while the other indicates the number of expected patterns. In this way, users do not need to struggle to specify a suitable and data-dependent value for the minimum support. Besides, users a reable to explicitly control the number of expected patterns. We experimentally studied the performance of the proposed fuzzy PSO and Ranking approach with different transaction datasets and various users’ expectation. The experimental results demonstrate the effectiveness and efficiency of the proposed fuzzy approach.