Fuzzy measurement based data processing method of k-means clustering

Abstract

Provided is a fuzzy measurement based data processing method of k-means clustering. The method comprises the following steps that 1) initialization is carried out, q vectors are selected from m n-dimensional vectors randomly and serve as an initial mean-value clustering center, and q represents the number of types; 2) the fuzzy measurement distance to each mean-value center of each vector to be clustered is calculated; 3) a type number is distributed to the vector to be clustered, and the type number is determined by the type of the mean-value center of the minimal fuzzy measurement distance; 4) vectors to be clustered are traversed, according to the type numbers, an average vector is calculated for the vectors with the same type numbers, and the average vector is updated as the new mean value center; 5) for each type, the fuzzy measurement distance between the present mean-value center and the mean-value center that is updated is calculated; and 6) if the fuzzy measurement distance between the mean-value centers is lower than a preset threshold, classification is completed, and otherwise, the step 2) is returned to. The data processing method satisfies subjective fuzzy standards more and is higher in accuracy.

Description

technical field [0001] The invention relates to the field of data processing, in particular to a data processing method of k-means clustering. Background technique [0002] The k-means clustering method is a widely used pattern recognition method, which can be applied to the pattern mining of data such as time series and digital images. The k-means clustering method uses the Euclidean distance to measure the difference between the two groups of data. By calculating the distance from the average vector of each category (note, vector and vector are the same term, there will be no distinction below, and they are mixed at any time), and the categories are assigned continuously. Repeat this process repeatedly until a stable category assignment result is obtained. This method iterates the process of calculating the average vector and assigning categories. Generally speaking, it can only guarantee to obtain locally optimal classification results in theory. The k-means clustering ...

AI Technical Summary

Problems solved by technology

[0007] In order to overcome the shortcomings of the poor accuracy of existing k-means clustering data processing methods, the present invention provides a fuzzy metric-based k-means clustering data processing that is more in line with human subjective fuzziness standards and has better accuracy method

Images