Workshop on Applications-Driven Geometric Functional Data Analysis

该奖项支持参加于2017年10月8日至10日在佛罗里达州塔拉哈西佛罗里达州立大学校园举行的为期三天的应用驱动几何功能数据分析（FDA）研讨会。研讨会将重点讨论FDA、无限维微分几何和数据驱动应用的交叉点，并将邀请来自这些领域的高级专家和初级研究人员。研讨会包括半天的教程，提供研讨会主题的介绍，特别是针对初级研究人员。剩下的时间将分配给邀请演讲、海报会议和讨论会议。这次研讨会的主要动机是发起和促进具有不同背景和专业知识的美国科学家和数学家之间的讨论和未来的合作。组织者将在研讨会的三个组成部分——几何、FDA和大数据分析——瞄准领先的专家，力求广度和多样性。海报环节将用于鼓励所有参与者，特别是初级研究人员，展示他们的研究并与他人互动。从自然科学、社会科学到工程技术，各个学科的功能数据都在快速增长，这使得FDA成为一个重要的研究课题。为了应对FDA面临的各种挑战，需要利用功能数据中存在的低维模式和结构。几何提供了工具：(1)提取和分析功能数据中的结构；(2)处理函数空间的无限维数；(3)得出可行的统计模型并进行有效的推论；(4)处理与许多当前和未来以数据为中心的应用程序相关的大型数据集。研讨会将包括：(1)理论主题，如无限维空间的几何，不变度量下的分析，以及涉及函数数据的统计模型；(2)应用，如计算解剖学、生物信息学、生物识别学和神经科学。这些重要和互补的主题使研讨会成为一个独特而有价值的活动。有关研讨会的详情，请浏览https://ani.stat.fsu.edu/GFDW/

This award supports participation in a three-day workshop on applications-driven geometric functional data analysis (FDA) held October 8-10, 2017 on the campus of Florida State University in Tallahassee, Florida. The workshop focuses on topics in the intersections of FDA, infinite-dimensional differential geometry, and data-driven applications, and will feature both senior experts and junior researchers from these areas. The workshop includes a half-day of tutorials providing an introduction to the workshop topics, especially aimed at junior researchers. The remaining time will be allocated to invited talks, a poster session, and discussion sessions. The main motivation for this workshop is to initiate and facilitate discussions and future collaborations between US scientists and mathematicians with different backgrounds and expertise. The organizers will strive for breadth and diversity by targeting leading experts in the three components areas of the workshop -- geometry, FDA, and big data analysis. The poster session will be used to encourage all participants, especially the junior researchers, to present their research and interact with others. A rapid growth of functional data in all disciplines, ranging from natural and social sciences to engineering and technology, has made FDA an important topic of research. To handle the diverse challenges facing FDA, there is a need to exploit low-dimensional patterns and structures present in functional data. Geometry provides tools to: (1) extract and analyze structures in functional data, (2) handle the infinite-dimensionality of function spaces, (3) arrive at viable statistical models and make efficient inferences; and (4) process large datasets associated with many current and future data-centric applications. The workshop will include both: (1) theoretical topics, such as geometries of infinite-dimensional spaces, analysis under invariant metrics, and statistical models involving functional data; and (2) applications, such as computational anatomy, bioinformatics, biometrics, and neuroscience. This blend of important and complementary topics makes the workshop a unique and worthwhile event. For more information about the workshop, please refer to https://ani.stat.fsu.edu/GFDW/