Distributed support vector clustering method and system

Abstract

The invention discloses a distributed support vector clustering method which comprises the following steps: processing the input data set according to a preset processing rule, and initializing global parameters and tasks; distributing a preset data set or a specific calculation result to each computational node; initializing a weight vector of the preset data set when the preset data set is distributed to the computational node; performing iterative operation according to a preset formula, and calculating a weight coefficient value of each sample in the preset data set; finding a sample of which the weight coefficient value is higher than a preset minimal value to serve as a support vector, and numbering the support vector; and building a support function by utilizing the support vector and the weight coefficient of each support vector, performing cluster division to obtain a cluster label of the support vector, and calibrating a non-support vector sample as a clustering analysis result. According to the method, the support vector clustering efficiency can be effectively improved.

Description

technical field [0001] The invention relates to the technical field of data processing, in particular to a distributed support vector clustering method and system. Background technique [0002] Clustering analysis is to find the observation value clusters with close relationship in the object set through some similarity measure, so that the similarity between the objects in the cluster is as large as possible, and the similarity between the objects of different clusters is as large as possible. Potentially small, even different or irrelevant. [0003] At present, the advantages and disadvantages of clustering analysis methods are carried out by measuring effectiveness and realizing performance, namely time efficiency and storage efficiency. [0004] The support vector clustering method is better for effectiveness. Among them, support vector clustering is a kind of method based on kernel function. It maps data from input space to high-dimensional feature space by using kerne...

AI Technical Summary

Problems solved by technology

However, the biggest disadvantage of this method is the poor performance, the huge consumption of storage space by the kernel matrix built on the entire training set, and the expensive time cost when solving the support function describing the hypersphere.

[0005] The existing methods for improving the efficiency of support vector clustering analysis mainly include 1) conversion to solve the dual problem of the support function, but although it is beneficial to the calculation efficiency, it does not help to reduce the size of the kernel matrix at one time; 2) training set reduction , but as an improvement in the preprocessing stage, the performance improvement of large-scale or high-dimensional (or both) data analysis is limited, and it is easy to introduce more parameters to increase the complexity of the algorithm

Therefore, none of these methods can effectively improve the efficiency of the support vector clustering analysis method.

Images