Data design and analysis based on cloud computing and improved K-Means algorithm

Issue title: Special Section: Intelligent Algorithms for Complex Information Services - Recent Advances and Future Trends

Guest editors: Andino Maseleno, Xiaohui Yuan and Valentina E. Balas

Article type: Research Article

Authors: Wu, Chunqiong^{a; b; *} | Yu, Rongrui^{a; b} | Yan, Bingwen^{a; b} | Huang, Zhangshu^{a; b} | Yu, Baoqin^{a; b} | Yu, Yanliang^{a; b} | Chen, Na^{b; c} | Zhou, Xiukao^{a; b}

Affiliations: [a] Business College, Yango University, Fujian Province, China | [b] Big Data Business Intelligence Engineering Research Center of Fujian University, Fujian, China | [c] Software Engineering Institute, Lanzhou Institute Technology, Gansu Province, China

Correspondence: [*] Corresponding author. Chunqiong Wu, E-mail: cqwu@ygu.edu.cn.

Abstract: The IoT and Artificial intelligence, the amount of information generated on the Web site is increasing. The rise of the Hadoop distributed cloud computing platform (HDCCP) makes it possible to use multiple computing nodes for parallel computing to solve the performance problems of traditional serial algorithms. The purpose of this paper is to study data design based on cloud computing and improved k-means algorithm (KMA). This paper deeply researches Hadoop distributed cloud computing platform and clustering algorithm and other related technologies, and designs and implements a cluster analysis system (CAS) based on HP. And through an in-depth analysis of the problems existing in the KMA, an improved scheme based on the HDP is designed. The experimental environment was conFig.d with the cluster analysis system implemented. Finally, the improved KMPA was tested experimentally from four directions: convergence speed, acceleration ratio, initialization sampling rate, and accuracy rate. We can see the experimental results that the CAS based on the HDCCP designed in this paper can provide efficient and configurable cluster analysis services. In this paper, the correct rate is 90.7%.

Keywords: Cloud computing, k-Means algorithm, cluster analysis, hadoop platform, data design

DOI: 10.3233/JIFS-179992

Journal: Journal of Intelligent & Fuzzy Systems, vol. 39, no. 4, pp. 5067-5074, 2020