0


从数据洪流中提取知识,揭示宇宙奥秘

Extracting Knowledge from the Data Deluge to Reveal the Mysteries of the Universe
课程网址: http://videolectures.net/iswc2019_johnston_hollitt_universe/  
主讲教师: Melanie Johnston-Hollitt
开课单位: 奇点大学
开课时间: 2019-12-10
课程语种: 英语
中文简介:

天体物理学是现代世界中数据最密集的研究领域之一,因此,它提供了独特的背景来推动“大数据”体系中的许多必要创新。尤其是射电天文学在大数据生成方面处于领先地位,得益于过去十年对该学科的持续全球投资,目前的望远镜每年生成数十 PB 的数据。这种所谓的“无线电复兴”的顶峰将是平方公里阵列(SKA)——一个全球天文台,其任务是探索宇宙最深的奥秘。 SKA 将创建有史以来最高分辨率、最快帧速率的进化宇宙电影,并且这样做将每天生成 160 TB 的原始数据,或每年接近 5 ZB 的数据。这些数据将被处理成每天 1 PB 的图像立方体,这些立方体将被处理、策划并最终通过协调的分层计算设施网络分发给全球天文学界进行科学开发。然而,这种真正数据丰富的环境将需要新的自动化和语义流程来充分利用生成的海量结果。事实上,要充分发挥这一努力的巨大科学潜力,我们不仅需要更好的数据标记和协调机制,还需要改进算法、人工智能、语义和本体,以自动方式跟踪和提取知识,规模不超过却在科学上有所尝试。在本主题演讲中,我将概述 SKA 项目,概述该项目面临的“大数据”挑战,并讨论我们为应对这一天文数据洪流而采取的一些方法。

课程简介: Astrophysics is one of the most data intensive research fields of the modern world and, as such, provides a unique context to drive many of the required innovations in the “big data” regime. In particular, radio astronomy is foremost in the field in terms of big data generation, and thanks to sustained global investment in the discipline over the last decade, present telescopes generate tens of petabytes of data per annum. The pinnacle of this so-called on-going ‘radio renaissance’ will be the Square Kilometre Array (SKA) — a global observatory tasked with probing the deepest mysteries of the Universe. The SKA will create the highest resolution, fastest frame rate movie of the evolving Universe ever and in doing so will generate 160 terrabytes of raw data a day, or close to 5 zettabytes of data per annum. These data will be processed into of order 1 petabyte of image cubes per day which will be processed, curated, and ultimately distributed via a network of coordinated tiered compute facilities to the global astronomical community for scientific exploitation. However, this truly data-rich environment will require new automated and semantic processes to fully exploit the vast sea of results generated. In fact, to fully realize the enormous scientific potential of this endeavour, we will need not only better data tagging and coordination mechanisms, but also improved algorithms, artificial intelligence, semantics and ontologies to track and extract knowledge in an automated way at a scale not yet attempted in science. In this keynote I will present an overview of the SKA project, outline the “big data” challenges the project faces and discuss some of the approaches we are taking to tame this astronomical data deluge.
关 键 词: 射电天文学; 平方公里阵列; 数据洪流
课程来源: 视频讲座网
数据采集: 2021-06-18:yumf
最后编审: 2021-06-18:yumf
阅读次数: 75