HDR TRIPODS: Institute for the Foundations of Graph and Deep Learning

HDR TRIPODS：图形和深度学习基础研究所

• 批准号: 1934979
• 负责人: Rene Vidal
• 金额: $ 150万
• 依托单位: Johns Hopkins University
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2019
• 资助国家: 美国
• 起止时间: 2019-10-01 至 2023-09-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1934979&HistoricalAwards=false
• 关键词: HDR; TRIPODS; Institute; Foundations; Graph

Classical data-analysis methods were based on mathematical, physical or statistical models for the data-generation process, which were developed under the assumption that the data were relatively clean and collected for a specific task. Over the past few decades, advances in data acquisition have led to massive, noisy, high-dimensional datasets, which were not necessarily collected for a specific task. This has lead to the emergence of data-driven methods, such as deep learning, which use massive amounts of labeled data to learn 'black-box' models, which do not provide an explicit description of the process being modeled. Such data-driven methods have led to dramatic improvements in the performance of pattern-recognition systems for applications in computer vision and speech recognition for which massive amounts of labeled data can be generated. However, existing models are not very interpretable, and their predictions are not robust to adversarial perturbations. Moreover, there are many applications in science and engineering where data labeling is extremely costly, and the ability to interpret model predictions and produce estimates of uncertainty is essential. To address these challenges, a TRIPODS Institute on the Theoretical Foundations of Data Science will be created at Johns Hopkins University. The goals of the institute will be to (1) develop the foundations for the next generation of data analysis methods, which will integrate model-based and data-driven approaches, (2) foster interactions among data scientists through a monthly seminar series, semester-long research themes, an annual research symposium, and a summer research school and workshop on the foundations of data science, and (3) create new undergraduate and graduate curricula on the foundations of data science.The institute brings together a multidisciplinary team of mathematicians, statisticians, theoretical computer scientists, and electrical engineers with expertise in the foundations of machine learning, deep learning, statistical learning and inference on graphs, optimization, approximation theory, signal processing, dynamical systems and controls, to develop the foundations for the next generation of data-analysis methods, which will integrate model-based and data-driven approaches. In particular, the institute will focus on studying the foundations of deep neural models (e.g., feedforward networks, recurrent networks, generative adversarial networks) and generative models of structured data (e.g., graphical models, random graphs, dynamical systems), with the ultimate goal of arriving at integrated models that are more interpretable, robust to perturbations, and learnable with minimal supervision. The goals of the Phase I Institute will be to (1) study generalization, optimization and approximation properties of feedforward networks, (2) develop the foundations of statistical inference and learning on and of graphs, and (3) study the integration of deep networks and graphs for learning maps between structured datasets. This project is part of the National Science Foundation's Harnessing the Data Revolution (HDR) Big Idea activity.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.

经典的数据分析方法基于数​​据生成过程的数学、物理或统计模型，这些模型是在数据相对干净并为特定任务收集的假设下开发的。在过去的几十年里，数据采集的进步带来了海量、嘈杂、高维的数据集，这些数据集不一定是为了特定任务而收集的。这导致了数据驱动方法的出现，例如深度学习，它使用大量标记数据来学习“黑盒”模型，而这些模型不提供对建模过程的明确描述。这种数据驱动的方法极大地提高了计算机视觉和语音识别应用中的模式识别系统的性能，可以生成大量的标记数据。然而，现有模型的解释性较差，并且它们的预测对于对抗性扰动并不稳健。此外，在科学和工程领域的许多应用中，数据标记的成本极其高昂，因此解释模型预测和产生不确定性估计的能力至关重要。为了应对这些挑战，约翰·霍普金斯大学将创建数据科学理论基础 TRIPODS 研究所。该研究所的目标是（1）为下一代数据分析方法奠定基础，该方法将整合基于模型和数据驱动的方法，（2）通过每月一次的研讨会系列、学期促进数据科学家之间的互动-长期研究主题、年度研究研讨会、暑期研究学校和数据科学基础研讨会，以及 (3) 在数据科学基础上创建新的本科生和研究生课程。该研究所汇集了一支多学科的数学家团队,统计学家、理论计算机科学家和电气工程师，拥有机器学习、深度学习、统计学习和图形推理、优化、逼近理论、信号处理、动力系统和控制基础知识的专业知识，为下一代数据分析方法，它将整合基于模型和数据驱动的方法。特别是，该研究所将重点研究深度神经模型（例如前馈网络、循环网络、生成对抗网络）和结构化数据生成模型（例如图模型、随机图、动态系统）的基础，并最终获得目标是获得更可解释、对扰动更鲁棒、并且可以在最少监督的情况下学习的集成模型。第一阶段研究所的目标是（1）研究前馈网络的泛化、优化和逼近特性，（2）开发统计推断和图学习的基础，以及（3）研究深度网络的集成以及用于学习结构化数据集之间的映射的图表。 该项目是美国国家科学基金会利用数据革命 (HDR) 大创意活动的一部分。该奖项反映了 NSF 的法定使命，并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，被认为值得支持。

项目成果

A Nullspace Property for Subspace-Preserving Recovery

用于保留子空间恢复的零空间性质

• 发表时间: 2021-07
• 期刊: Proceedings of Machine Learning Research
• 作者: Kaba, Mustafa Devrim;You, Chong;Robinson, Daniel R;Mallada, Enrique;Vidal, Rene
• 通讯作者: Vidal, Rene

Learning coherent clusters in weakly‑connected network systems

学习弱连接网络系统中的相干集群

• 发表时间: 2023-07
• 期刊: Proceedings of The 5th Annual Learning for Dynamics and Control Conference
• 作者: Min, Hancheng;Mallada, Enrique
• 通讯作者: Mallada, Enrique

Linear-Convex Optimal Steady-State Control

线性凸最优稳态控制

• DOI: 10.1109/tac.2020.3044275
• 发表时间: 2021-11
• 期刊: IEEE Transactions on Automatic Control
• 影响因子: 6.8
• 作者: Lawrence, Liam S.;Simpson;Mallada, Enrique
• 通讯作者: Mallada, Enrique

Accurate Reduced-Order Models for Heterogeneous Coherent Generators

异构相干发生器的精确降阶模型

• DOI: 10.1109/lcsys.2020.3043733
• 发表时间: 2021-11
• 期刊: IEEE Control Systems Letters
• 影响因子: 3
• 作者: Min, Hancheng;Paganini, Fernando;Mallada, Enrique
• 通讯作者: Mallada, Enrique

Spectral clustering and model reduction for weakly-connected coherent network systems

弱连接相干网络系统的谱聚类和模型简化

• DOI: 10.23919/acc55779.2023.10156212
• 发表时间: 2023-05
• 期刊: American Control Conference
• 作者: Min, Hancheng;Mallada, Enrique
• 通讯作者: Mallada, Enrique