Density peak-core fusion-based adaptive clustering method

Abstract

The invention provides a density peak-core fusion-based adaptive clustering method, and belongs to the field of pattern recognition. The method comprises the following steps: (1) firstly, carrying outdensity neighbor clustering based on a density peak value: calculating the density of each data point in the to-be-clustered data set by using a core density-k-nearest neighbor density estimation method, determining density peak points through an adaptive threshold value, and performing density nearest neighbor clustering by taking the density peak points as a class center to obtain an initial clustering result; And (2) performing core fusion operation based on intra-class divergence: adaptively determining a core point in each initial class, fusing the two initial classes if the core pointsof the two initial classes are adjacent and the intra-class divergence can be reduced after fusion, and fusing all the initial classes to be fused to obtain a final clustering result. The method is simple, accurate and efficient, and can perform self-adaptive clustering on the data set with any shape and density distribution.

Description

technical field [0001] The invention relates to an adaptive clustering method based on density peak-core fusion, which belongs to the field of pattern recognition. Background technique [0002] Clustering method is an important technology in the field of pattern recognition and machine learning, and is widely used in face recognition, search engineering, image partitioning and other fields. Clustering is the process of dividing a data set into classes or clusters according to the similarity between data points. Data points belonging to the same class have a greater similarity, while data points belonging to different classes should be as dissimilar as possible. . The similarity between data points can be measured by distance, and the most common distance is Euclidean distance. Due to the diversity of data sources, properties, and distributions, as well as the needs of automated industrial processes, clustering methods are required in many fields to automatically determine ...

AI Technical Summary

Problems solved by technology

Although the CFSFDP method performs adaptive clustering on data sets with arbitrary shape distributions, this method also has some shortcomings.

First, the clustering effect of the CFSFDP method is easily affected by the density estimation results

Secondly, the process of artificially determining the class center in the CFSFDP method limits its application effect in automated tasks

The most important point is that there may be multiple density peak points in a class. The CFSFDP method will regard multiple density peak points in the class as the class center, and a class with multiple density peak points will be mistakenly divided into multiple classes. , so the CFSFDP method cannot ensure correct clustering results

[0005] To sum up, although some existing clustering methods can adaptively process data sets with arbitrary shape distributions, the shape and density of these data sets are mostly presented as simple distributions. Accuracy and efficiency still need to be improved

Images