A Virtual Conference on Applications of Statistical Methods and Machine Learning in the Space Sciences; Boulder, Colorado; March 22-26, 2021 or June 28 - July 2, 2021

该奖项是为了支持一个虚拟会议，统计方法和机器学习在空间科学中的应用，旨在汇集学术界和工业界的专家，利用统计学，数据科学，人工智能（AI）和信息理论的最新进展，利用空间科学领域的大量数据集。 会议预计将有来自空间科学和数据科学家，统计学家和人工智能专家的所有学科的研究人员参加。会议将是一个独特的机会，分享探索“大数据”的最新趋势和进展，并促进跨学科合作。机器学习是空间科学的一个新兴趋势，目的是从使用航天器和地面观测获得的大量数据中识别模式和提取信息。统计方法已经在数据分析中使用了几十年，这次会议的目的是将这两种技术结合起来分析空间科学中的“大数据”。机器学习（ML）算法，特别是神经网络，为应用特定方法的问题提供了一个“黑箱”解决方案。有一些先进的统计方法可以单独使用，也可以与机器学习算法结合使用，以更深入地探索ML模型中输入和输出变量之间的关系，以及底层物理学。 会议将重点关注“黑盒”与“可解释”模型-即理解系统的物理和动力学，同时使用ML方法寻求准确的解决方案会议将为新研究人员提供使用这些多功能和新颖的数据分析工具的培训。最近的毕业生和早期职业研究人员将更好地了解跨学科方法的好处，以了解他们各自感兴趣的领域的基础科学问题。该奖项反映了NSF的法定使命，并已被认为是值得通过使用基金会的智力价值和更广泛的影响审查标准进行评估的支持。

The award is to support a virtual conference, Applications of Statistical Methods and Machine Learning in the Space Sciences, which aims to bring together experts in academia and industry to leverage recent advancements in statistics, data science, artificial intelligence (AI) and information theory to make use of large volume datasets in the field of space sciences. The conference anticipates participation of researchers from all disciplines of space science and data scientists, statisticians and AI experts. The conference will be a unique opportunity for sharing recent trends and advancements in exploring “big data”, and for fostering interdisciplinary collaborations. Machine learning is an emerging trend in the space sciences to identify patterns and extract information from the enormous data acquired using spacecraft and ground-based observations. Statistical methods have already been in use in data analysis for decades and the aim of the conference is to bring these two techniques together in the analysis of “big data” in the space sciences. Machine learning (ML) algorithms, particularly neural networks, present a “black box” solution to the problem in which a particular method is applied. There are advanced statistical methods that can be used either by themselves or in conjunction with a machine learning algorithm to probe deeper into the relationships between input and output variables in the ML model, and the underlying physics. The conference will have an emphasis on “black box” versus “interpretable” models — that is, on understanding the physics and dynamics of the system while seeking an accurate solution using ML methodsThe conference will provide training for new researchers in the use of these versatile and novel tools of data analysis. Recent graduates and early career researchers will be better equipped with an appreciation of the benefits of interdisciplinary approaches to understanding the fundamental science problems in their respective fields of interest.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.