Frequent pattern mining is a basic problem as well as an important task in data mining and knowledge discovery. But however, frequent patterns in large scale datasets is an extremely time consuming task. So, in order to reduce the time the fuzzy concept was introduced in order to discretize the dataset in the range of 0 to 1. The Particle Swarm Optimizing (PSO) algorithm was basically developed from the social behavior of various animals likes bird flocking and fish schooling etc. In the existing versions of the local PSO algorithm with different neighborhood structures and the multi swarm PSOs, the swarms are predefined or dynamically adjusted according to the distance. Due to this the freedom of sub-swarms is limited. But in dynamic multi-swarm particle swarm optimizer (DMS-PSO) the neighborhood structure is dynamic and randomized. In this paper, a fuzzy data set has been used and various frequent pattern mining techniques like Apriori, Vertical data format and Frequent Pattern (FP) growth were implemented. Out of various frequent pattern mining techniques it was clear that FP growth method yields the better results on a fuzzy dataset. The frequent patterns obtained were considered as the set of initial population or particles. For the selection criteria, we have considered the mean squared residue (MSR) score rather using the threshold value. It has been observed that DMS-PSO based fuzzy FP growth technique finds the best individual frequent patterns as compared to the traditional PSO based fuzzy FP growth and also the runtime of the first was much better than the latter.