The Study of Improved FP-Growth Algorithm in MapReduce
Sun Hong, Zhang Huaxuan, Chen Shiping, Hu Chunyan · Advances in intelligent systems research/Advances in Intelligent Systems Research · 2013
As FP-Growth algorithm generates a great deal of conditional pattern bases and conditional pattern trees, leading to low efficiency, propose an improved FP-Growth(IFP) algorithm which firstly combine the same patterns based on the situation whether the support of the transaction is greater than the minimum support(min_sup) to mine the frequent patterns.Thus the IFP cuts down on the space and improves the efficiency.It also makes it easy to be paralleled.Further more, combine the IFP algorithm with the MapReduce computing model, named MR-IFP(MapReduce-Improved FP), to improve the capability to deal with the mass data.