[Retracted] Text Clustering and Economic Analysis of Free Trade Zone Governance Strategies Based on Random Matrix and Subject Analysis

Algorithm 2

Initial clustering algorithm.

	Input: the centroids of the input semantic clusters are used as the initial cluster centers.
	Output: the new vector values of each cluster member space transformed and the new vector values of the test data set are input into the machine.
	repeat.
	Calculating the similarity of each point to the cluster centers by the new vector value, forming n clusters.
	recalculating the center of mass of each cluster.
	until: stop after n clusters converge.