平行数据:从大数据到数据智能
发布时间：2019-06-09 16:50
【摘要】：为了解决实际问题,大数据分析处理系统需要获取数据,然而实际场景中收集到的实际数据通常不完备.另外,大多数问题的解决方案通常是由问题引导或者仅仅进行数据分析,运行参数调整和设定带有较大的盲目性,难以达到应用的智能性.为此,文中提出平行数据的概念和框架,根据实际数据经计算实验产生真正的虚拟大数据,结合默顿定律,以期待的解决方案与问题进行广义对偶,引导大数据聚焦到实际问题.实际数据与虚拟数据动态互动,平行演化,形成一个虚实相生、数据动态变化的过程,最终使数据具备智能,进而解决未知的问题.平行数据不但是一种数据表示形式,更是一种数据演化机制与方式,其特色是虚实互动,所有数据的动力学轨迹构成了数据动力学系统.平行数据为数据处理、表示、挖掘和应用提供了一个新的范式.
[Abstract]:In order to solve the practical problems, big data analysis and processing system needs to obtain data, but the actual data collected in the actual scene is usually incomplete. In addition, most of the solutions to the problem are usually guided by the problem or just for data analysis, and the adjustment and setting of the running parameters are blind, so it is difficult to achieve the intelligence of the application. For this reason, this paper puts forward the concept and framework of parallel data, produces real virtual big data according to the actual data through calculation and experiment, and combines Merton's law to carry on the generalized duality with the expected solution and the problem. Guide big data to focus on practical problems. Real data and virtual data dynamic interaction, parallel evolution, the formation of a virtual reality, the dynamic change of data process, and finally make the data intelligent, and then solve unknown problems. Parallel data is not only a data representation, but also a data evolution mechanism and mode, which is characterized by virtual and real interaction. The dynamic trajectories of all the data constitute the data dynamics system. Parallel data provides a new paradigm for data processing, representation, mining and application.
【作者单位】： 中国石油大学(华东)计算机与通信工程学院;中国科学院自动化研究所复杂系统管理与控制国家重点实验室;青岛智能产业技术研究院平行工作室;西安交通大学人工智能与机器人研究所;国防科学技术大学军事计算实验与平行系统技术研究中心;
【基金】：国家自然科学基金重点项目(No.61533019,71232006,61233001);国家自然科学基金项目(No.61702519,61309024) 科技部创新方法工作专项项目(No.2015IM010300) 中央高校基本科研业务费项目(No.2015020031) 中石油创新基金项目(No.2016D-5007-0305) 山东省自然科学基金项目(No.ZR2014FM038) 山东省重点研发计划(No.2017GGX10140)资助~~
【分类号】：TP311.13
，
本文编号：2495717
[Abstract]:In order to solve the practical problems, big data analysis and processing system needs to obtain data, but the actual data collected in the actual scene is usually incomplete. In addition, most of the solutions to the problem are usually guided by the problem or just for data analysis, and the adjustment and setting of the running parameters are blind, so it is difficult to achieve the intelligence of the application. For this reason, this paper puts forward the concept and framework of parallel data, produces real virtual big data according to the actual data through calculation and experiment, and combines Merton's law to carry on the generalized duality with the expected solution and the problem. Guide big data to focus on practical problems. Real data and virtual data dynamic interaction, parallel evolution, the formation of a virtual reality, the dynamic change of data process, and finally make the data intelligent, and then solve unknown problems. Parallel data is not only a data representation, but also a data evolution mechanism and mode, which is characterized by virtual and real interaction. The dynamic trajectories of all the data constitute the data dynamics system. Parallel data provides a new paradigm for data processing, representation, mining and application.
【作者单位】： 中国石油大学(华东)计算机与通信工程学院;中国科学院自动化研究所复杂系统管理与控制国家重点实验室;青岛智能产业技术研究院平行工作室;西安交通大学人工智能与机器人研究所;国防科学技术大学军事计算实验与平行系统技术研究中心;
【基金】：国家自然科学基金重点项目(No.61533019,71232006,61233001);国家自然科学基金项目(No.61702519,61309024) 科技部创新方法工作专项项目(No.2015IM010300) 中央高校基本科研业务费项目(No.2015020031) 中石油创新基金项目(No.2016D-5007-0305) 山东省自然科学基金项目(No.ZR2014FM038) 山东省重点研发计划(No.2017GGX10140)资助~~
【分类号】：TP311.13
，
本文编号：2495717
本文链接：https://www.wllwen.com/kejilunwen/ruanjiangongchenglunwen/2495717.html
