Adaptive Peak Environmental Density Clustering Algorithm in Cloud Computing Technology

Introduction

Clustering, which is widely used in pattern recognition, data mining and other fields, is a way to classify data sets without manual supervision. Clustering algorithm groups data according to similarity of data. With the development of intelligent cloud computing and network communication technology, cloud grid communication technology is utilized to data fusion scheduling and storage to improve data storage performance. In the environment of cloud technology, data is stored in the form of sparse grid, and the distribution of data is unbalanced, resulting in poor effective clustering detection and recognition ability of data. It is necessary in perform optimization detection and clustering processing on the grid sparse unbalanced cloud data. Fusion clustering analysis model of grid sparse unbalanced cloud data is built, and the clustering analysis and recognition ability of grid sparse unbalanced cloud data are improved by combining spatial clustering feature analysis and optimized data mining algorithm. Research on the clustering method of unbalanced grid sparsity cloud data has attracted great attention. About unbalanced grid sparse cloud clustering of data processing are mainly fuzzy C-means clustering, irregular triangle gathering originative and cloud data clustering method. The distributed structure model of grid sparse unbalanced cloud data was built, and cluster of grid sparse unbalanced cloud data was carried out through the reorganization of the spatial distributed structure. Reference (You, 2019) proposes the density peak clustering algorithm of grid sparse unbalanced cloud data environment built on deep learning. Heterogeneous directed graph fusion method is adopted to design the storage structure of grid sparse unbalanced cloud data set, and combined with feature space reorganization technology to carry out grid sparse unbalanced cloud data set structure reorganization, so as to improve the fusion property of data clustering. However, this method has a large computing overhead and poor real-time performance in data clustering. Reference (Ma, 2019) puts forward clustering method of grid sparse unbalanced cloud data based on fuzzy C-means clustering. Combined with grid shading peak clustering and attribute classification and recognition, deep learning method is adopted to carry out optimization learning of data clustering process.

Reference (Tang, Xinyu (2019)) is proposed based on swarm intelligence algorithm in cloud computing of large data mining, clustering analysis, clustering algorithm of fuzzy C - average clustering algorithm, the heuristic hybrid leapfrog algorithm of swarm intelligence optimization techniques combined with fuzzy C - average clustering, in order to adjust the parameters of the less optimization under the condition of global search ability, which can better solve the problem of local trap, with good clustering effect, accuracy and convergence speed. At the same time, the algorithm has high stability, but the fuzzy clustering analysis of this method is easy to fall into local extremum.

However, this method has low feature recognition for data clustering. To solve the above problems, this paper proposes an improved adaptive peak density clustering algorithm under cloud computing technology. Building uneven meshes thin cloud data storage structure model. Through the characteristics of the strict registration method adaptive under cloud computing environment of peak density feature extraction and matching, according to data set of distribution grid blocks for data fusion and peak feature of cloud cluster, extract uneven mesh thin cloud peak characteristic of data set. Through space spectrum characteristics of clustering and information fusion method, it carries on data of binary semantic characteristics of distributed detection to obtain adaptive environment under cloud computing technology density peak cluster. Finally, simulation test and analysis are carried out to demonstrate superior performance of the proposed method in improving adaptive peak density clustering ability of the environment under cloud computing technology.