Vector clustering method, device and storage medium

Abstract

The application discloses a vector clustering method, device and storage medium, belonging to the field of data processing. The method includes: dividing the N vectors to be clustered into M groups to obtain M vector groups, combining the M vector groups into M groups, each group including an A vector group, and one or more B vector group, determine the similarity between any two vectors in each A vector group, and the similarity between each vector included in the A vector group and each B vector group included in the same group , to obtain the similarity between every two vectors in the N vectors, and cluster the N vectors according to the similarity between every two vectors in the N vectors. The present application can avoid the problem of repeated calculation, thereby reducing calculation overhead, and does not occupy too much storage space.

Description

technical field

[0001] The present application relates to the field of data processing, in particular to a vector clustering method, device and storage medium. Background technique

[0002] With the development of machine learning, the characteristics of many entities in reality can be represented by vectors. In this case, when the entity needs to be classified, the vector used to represent the feature of the entity can be clustered, and the classification of the entity can be represented by the clustering of the vector.

[0003] In related technologies, a computer device may divide multiple vectors into multiple vector groups, where each vector group contains multiple vectors. Afterwards, for each vector group in the plurality of vector groups, the computer device can calculate the similarity between every two vectors in this vector group, and the similarity between each vector in this vector group and the remaining vector groups. The similarity between each vector, and t...

Images