Newton STFC-Narit: Using astronomy surveys to train Thai researchers in handling Big Data

Newton STFC-Narit：利用天文学调查来培训泰国研究人员处理大数据

• 批准号: ST/R006539/1
• 负责人: James Mullaney
• 金额: $ 11.38万
• 依托单位: University of Sheffield
• 依托单位国家: 英国
• 项目类别: Research Grant
• 财政年份: 2018
• 资助国家: 英国
• 起止时间: 2018 至 无数据
• 项目状态: 已结题
• 来源: https://gtr.ukri.org/projects?ref=ST%2FR006539%2F1
• 关键词: Newton; STFC; Narit; Using; astronomy

The most effective way of reducing levels of poverty in developing countries is through their continued economic development, which leads to increased levels of income per person. To remain competitive, however, a developing economy needs access to a workforce with increasingly sophisticated skills. For Thailand today, this means skills that enable innovation, allowing it to successfully compete against other developing and developed economies. With more and more sectors collecting data on their customers, production lines, distribution networks, stock prices, etc., one of the most crucially needed "high-level" skills is the ability to handle large amounts of digital data. However, Thai data scientists and students typically lack ready access to very large datasets, which presents a barrier to their training in this area. Similarly, other scientists - including astronomers - typically lack the necessary data handling skills to efficiently store and analyse the large amounts of data they collect. Our project addresses both these problems by combining UK and Thai astronomers' access to and understanding of very large datasets with Thai data scientists' skills in databasing and machine learning to train Thai students in advanced data handling techniques.Working under the supervision of the Thai and UK partners, the graduate students involved in the project will establish a data centre at NARIT to store and automatically analyse the hundreds of gigabytes of data generated each night by the Gravitational-wave Optical Transient Observatory (GOTO) - a major new survey telescope of which NARIT is a contributing member. In the process, the students will gain vital experience of database management and automated, machine learning-based data analyses. The resulting data centre will be an important research asset for NARIT astronomers and a key training resource for the broader Thai scientific community. Indeed, we will ourselves use the data centre as a training aid in teaching data handling skills to up to 60 other researchers and students during two 5-day practical workshops (one held each year of the grant with space for up to 30 trainees each). Through three graduate research projects, our team will develop the data centre into an automated storage and analysis system with the ultimate goal of outputting a prioritised list of targets for follow-up observations with NARIT's other observing facilities. In doing so, the Thai data centre will be a testbed for machine learning-based analyses, remaining at the forefront of all GOTO data centres in terms of data handling research. On completion of the project, the skills acquired by the Thai students will be readily transferrable to a diverse range of economic sectors such as information technology, medicine, finance, security, etc., thereby helping the further economic development of Thailand.

减少发展中国家贫穷程度的最有效办法是通过持续的经济发展，从而提高人均收入水平。然而，为了保持竞争力，发展中经济体需要获得技能日益复杂的劳动力。对今天的泰国来说，这意味着能够促进创新的技能，使其能够成功地与其他发展中国家和发达经济体竞争。随着越来越多的行业收集客户、生产线、分销网络、股票价格等数据，最迫切需要的“高级”技能之一是处理大量数字数据的能力。然而，泰国的数据科学家和学生通常无法随时访问非常大的数据集，这对他们在这一领域的培训构成了障碍。与此类似，包括天文学家在内的其他科学家通常缺乏必要的数据处理技能来有效地存储和分析他们收集的大量数据。我们的项目通过将英国和泰国天文学家对超大型数据集的访问和理解与泰国数据科学家在数据库和机器学习方面的技能相结合来解决这两个问题，以培训泰国学生掌握先进的数据处理技术。在泰国和英国合作伙伴的监督下，参与该项目的研究生将在NARIT建立一个数据中心，以存储和自动分析引力波光学瞬态观测台（后藤）每晚产生的数百千兆字节数据。在这个过程中，学生将获得数据库管理和自动化，基于机器学习的数据分析的重要经验。由此产生的数据中心将成为国家研究和信息技术研究所天文学家的重要研究资产，也是泰国广大科学界的关键培训资源。事实上，我们自己将使用数据中心作为培训辅助工具，在两个为期5天的实践研讨会期间向多达60名其他研究人员和学生教授数据处理技能（每年举办一次，每次最多可容纳30名学员）。通过三个研究生研究项目，我们的团队将把数据中心开发成一个自动化的存储和分析系统，最终目标是输出一个优先目标列表，以便与NARIT的其他观测设施进行后续观测。在此过程中，泰国数据中心将成为基于机器学习的分析的试验台，在数据处理研究方面保持在所有后藤数据中心的最前沿。在项目完成后，泰国学生获得的技能将很容易转移到各种经济部门，如信息技术，医药，金融，安全等，从而促进泰国经济的进一步发展。

项目成果

Handling Imbalance Problem in Convolutional Neural Network for Astronomical Data Classification

天文数据分类中卷积神经网络不平衡问题的处理

• 发表时间: 2019
• 作者: Liu, J.J.

Transient Detection Modeling as Imbalance Data Classification

• DOI: 10.1109/ickii.2018.8569123
• 发表时间: 2018-07
• 期刊: 2018 1st IEEE International Conference on Knowledge Innovation and Invention (ICKII)
• 作者: Aireen B. Tabacolde;Tossapon Boongoen;Natthakan Iam-on;J. Mullaney;U. Sawangwit;K. Ulaczyk