Clustering result evaluation method and device based on large-scale samples

Abstract

The embodiment of the invention discloses a clustering result evaluation method and device based on large-scale samples. The embodiment of the invention provides a technical scheme. The method comprises: obtaining clustering results, randomly extracting a first set number of classes from all the classes of the clustering result, taking the clustering result as a first sampling class, extracting asecond set number of samples from each class of the first sampling class according to a set sampling rule to serve as sampling samples, forming a second sampling class based on the sampling samples, finally calculating a contour coefficient according to the second sampling class, and obtaining corresponding clustering result evaluation. By adopting the technical means, the data volume of the clustering result can be reduced through reasonable sample sampling, the complexity of clustering result evaluation calculation is reduced on the premise of ensuring that the sampled sample has the representativeness of the clustering result, and the generation efficiency of clustering result evaluation is further improved.

Description

technical field [0001] The embodiments of the present application relate to the technical field of data clustering, in particular to a method and device for evaluating clustering results based on large-scale samples. Background technique [0002] Clustering is a method of data mining. In the field of image processing technology, image clustering is a process of dividing multiple images into multiple classes composed of similar images based on image features. Image clustering plays an important role in image segmentation, object tracking and other fields. In many clustering scenarios, errors may occur in the clustering process. If images that do not belong to the same class are classified into the same image cluster, the center of the image cluster will shift, making the subsequent clustering results more accurate. increasingly inaccurate. Therefore, it is necessary to evaluate the accuracy of the clustering results and correct the wrong image clustering results in time. ...

AI Technical Summary

Problems solved by technology

However, when using the silhouette coefficient to evaluate the clustering results of large-scale samples, due to the relatively large number of samples to be evaluated, the computational complexity of the clustering result evaluation based on the silhouette coefficient is relatively high, and the generation efficiency of the clustering result evaluation is relatively low. lower

Images