粒计算:一种大数据融合智能建模新方法  

Granular computing: a new method of intelligent modeling for big data fusion

作　　者：姚富光 钟先信[2] 周靖超 Yao Fuguang;Zhong Xianxin;Zhou Jingchao(Information Center,Chongqing University of Education,Chongqing 400065,China;Key Laboratory of Optoelectronics Technology and Systems of the Education Ministry of China,Chongqing University,Chongqing 400044,China;School of Engineering,Rutgers University,New Jersey,08901 USA)

机构地区：[1]重庆第二师范学院信息中心,重庆400065 [2]重庆大学光电技术及系统教育部重点实验室,重庆400044 [3]罗格斯大学工程学院,美国新泽西08901

出　　处：《南京理工大学学报》2018年第4期503-510,共8页Journal of Nanjing University of Science and Technology

基　　金：教育部学校规划建设发展中心课题(CSDP18FC3204);重庆市教育规划课题(2017-GX-139);重庆第二师范学院科技项目(KY2016TZ02)

摘　　要：大数据融合智能是数据科学与工程的一个核心问题。大数据时代,深度学习新一代人工智能技术运用于复杂系统领域,面临全新的机遇和技术挑战。首先分析了大数据融合问题的内涵和特点,然后从复杂系统角度分析了深度学习技术研究新视角和粒计算在信息融合方面的新技术动态,推测了粒计算与深度学习的融合拓展可行性。最后,探讨了面向复杂系统认知研究的大数据融合智能建模粒计算处理架构,力求为复杂系统管理与控制技术研究拓展方法思路。Big data fusion intelligence is a core problem of data science and engineering.In the era of big data,as new generation of artificial intelligence technology,deep learning is applied to the field of complex systems,facing new opportunities and serious technical challenges.This paper first analyzes the connotation of the problem of big data fusion.The new visual angle of deep learning technology research and the new technology dynamics of granular computing in information fusion are analyzed from the point of view of complex system.The feasibility of the fusion of granular computing and deep learning is deduced.Finally,the architecture of large data fusion intelligent modeling particle computing and processing for complex system cognition research is discussed.The research is a meaningful work for development method of the system management and control research.

关 键 词：复杂系统 大数据融合 粒计算 深度学习 人工智能 认知计算 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]