Optimization-based Implicit Deep Learning, Theory and Applications
基于优化的隐式深度学习、理论与应用
基本信息
- 批准号:2309810
- 负责人:
- 金额:$ 29.5万
- 依托单位:
- 依托单位国家:美国
- 项目类别:Continuing Grant
- 财政年份:2023
- 资助国家:美国
- 起止时间:2023-07-15 至 2026-06-30
- 项目状态:未结题
- 来源:
- 关键词:
项目摘要
The past decade has seen remarkable success in deep learning. However, a significant challenge in today's era is to ensure interpretability and reliability in these models. In various applications, deep neural networks (DNNs) need to provide guarantees on their outputs, such as maintaining a self-driving car within its lane. On the other hand, many of these tasks can be formulated as optimization problems, where optimization algorithms offer interpretable and reliable solutions. Unfortunately, these models do not leverage data and thus fall short of state-of-the-art deep learning models. This research will address enhancing interpretability and reliability in deep learning methods and improve public safety when such learning methods are applied. In addition, the project will provide valuable educational opportunities for students involved. Participants will gain knowledge in inverse problems, optimization, and machine learning, which are transferable skills applicable in academia, government, and industry. The project aims to develop a framework that combines the interpretability and reliability of optimization algorithms with the design and training of DNNs. The primary focus is on implicit networks, a type of DNNs that determines their outputs implicitly through fixed point or optimality conditions, rather than a fixed number of computations like traditional DNNs with a set number of layers. This integration of optimization algorithms into implicit networks is referred to as implicit learning-to-optimize (L2O) networks. Implicit L2O networks have the potential to overcome the limitations of traditional DNNs, including their lack of reliability and interpretability. However, training and designing implicit L2O models present additional challenges that hinder their widespread adoption. To address these challenges, the research aims to develop a universal implicit L2O framework.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.
过去十年在深度学习中取得了巨大的成功。但是,在当今时代的一个重大挑战是确保这些模型中的可解释性和可靠性。在各种应用中,深层神经网络(DNN)需要为其产出提供保证,例如在其车道内保持自动驾驶汽车。另一方面,其中许多任务可以作为优化问题提出,优化算法提供了可解释且可靠的解决方案。不幸的是,这些模型不利用数据,因此缺乏最新的深度学习模型。这项研究将解决深度学习方法中的可解释性和可靠性,并在应用这种学习方法时提高公共安全。此外,该项目将为参与学生提供宝贵的教育机会。参与者将获得相反问题,优化和机器学习的知识,这些技能适用于学术界,政府和行业。该项目旨在开发一个将优化算法的可解释性和可靠性与DNN的设计和培训相结合的框架。主要的重点是隐式网络,一种DNN的类型,它通过固定点或最佳条件暗中确定其输出,而不是固定数量的计算,例如传统的DNN,具有一定的层。将优化算法集成到隐式网络中的这种集成被称为隐式学习到优势(L2O)网络。隐式L2O网络有可能克服传统DNN的局限性,包括缺乏可靠性和解释性。但是,培训和设计隐式L2O模型提出了其他挑战,从而阻碍了他们广泛采用。为了应对这些挑战,该研究旨在开发一个普遍的隐式L2O框架。该奖项反映了NSF的法定任务,并被认为是值得通过基金会的知识分子优点和更广泛的影响评估标准来评估的。
项目成果
期刊论文数量(0)
专著数量(0)
科研奖励数量(0)
会议论文数量(0)
专利数量(0)
数据更新时间:{{ journalArticles.updateTime }}
{{
item.title }}
{{ item.translation_title }}
- DOI:
{{ item.doi }} - 发表时间:
{{ item.publish_year }} - 期刊:
- 影响因子:{{ item.factor }}
- 作者:
{{ item.authors }} - 通讯作者:
{{ item.author }}
数据更新时间:{{ journalArticles.updateTime }}
{{ item.title }}
- 作者:
{{ item.author }}
数据更新时间:{{ monograph.updateTime }}
{{ item.title }}
- 作者:
{{ item.author }}
数据更新时间:{{ sciAawards.updateTime }}
{{ item.title }}
- 作者:
{{ item.author }}
数据更新时间:{{ conferencePapers.updateTime }}
{{ item.title }}
- 作者:
{{ item.author }}
数据更新时间:{{ patent.updateTime }}
Samy Wu Fung其他文献
Structured World Representations in Maze-Solving Transformers
迷宫解决变压器中的结构化世界表示
- DOI:
- 发表时间:
2023 - 期刊:
- 影响因子:0
- 作者:
Michael I. Ivanitskiy;Alex F Spies;Tilman Rauker;Guillaume Corlouer;Chris Mathwin;Lucia Quirke;Can Rager;Rusheb Shah;Dan Valentine;Cecilia G. Diniz Behn;Katsumi Inoue;Samy Wu Fung - 通讯作者:
Samy Wu Fung
A Neural Network Approach for High-Dimensional Optimal Control Applied to Multiagent Path Finding
应用于多智能体路径查找的高维最优控制神经网络方法
- DOI:
10.1109/tcst.2022.3172872 - 发表时间:
2021 - 期刊:
- 影响因子:4.8
- 作者:
Derek Onken;L. Nurbekyan;Xingjian Li;Samy Wu Fung;S. Osher;Lars Ruthotto - 通讯作者:
Lars Ruthotto
Global Solutions to Nonconvex Problems by Evolution of Hamilton-Jacobi PDEs
Hamilton-Jacobi 偏微分方程演化的非凸问题全局解
- DOI:
10.1007/s42967-022-00239-5 - 发表时间:
2022 - 期刊:
- 影响因子:1.6
- 作者:
Howard Heaton;Samy Wu Fung;S. Osher - 通讯作者:
S. Osher
Faster Predict-and-Optimize with Three-Operator Splitting
通过三算子分割加快预测和优化速度
- DOI:
10.48550/arxiv.2301.13395 - 发表时间:
2023 - 期刊:
- 影响因子:0
- 作者:
Daniel Mckenzie;Samy Wu Fung;Howard Heaton - 通讯作者:
Howard Heaton
A multiscale method for model order reduction in PDE parameter estimation
偏微分方程参数估计中模型降阶的多尺度方法
- DOI:
- 发表时间:
2017 - 期刊:
- 影响因子:2.4
- 作者:
Samy Wu Fung;Lars Ruthotto - 通讯作者:
Lars Ruthotto
Samy Wu Fung的其他文献
{{
item.title }}
{{ item.translation_title }}
- DOI:
{{ item.doi }} - 发表时间:
{{ item.publish_year }} - 期刊:
- 影响因子:{{ item.factor }}
- 作者:
{{ item.authors }} - 通讯作者:
{{ item.author }}
{{ truncateString('Samy Wu Fung', 18)}}的其他基金
Development of Geometrically-Flexible Physics-Based Convolution Kernels
基于几何灵活物理的卷积核的开发
- 批准号:
2110745 - 财政年份:2021
- 资助金额:
$ 29.5万 - 项目类别:
Standard Grant
相似国自然基金
基于隐式几何描述的梯度随机点阵结构多尺度建模与优化设计
- 批准号:12372200
- 批准年份:2023
- 资助金额:52 万元
- 项目类别:面上项目
基于七阶隐式大涡模拟的内外流耦合抽吸流动控制机理与优化研究
- 批准号:12202297
- 批准年份:2022
- 资助金额:30.00 万元
- 项目类别:青年科学基金项目
基于七阶隐式大涡模拟的内外流耦合抽吸流动控制机理与优化研究
- 批准号:
- 批准年份:2022
- 资助金额:30 万元
- 项目类别:青年科学基金项目
公钥密码的格分析优化方法研究
- 批准号:61702505
- 批准年份:2017
- 资助金额:26.0 万元
- 项目类别:青年科学基金项目
基于光滑变形隐式曲线/曲面的结构拓扑优化方法研究
- 批准号:11702254
- 批准年份:2017
- 资助金额:25.0 万元
- 项目类别:青年科学基金项目
相似海外基金
Spatial Calibration of Head-Mounted Displays Based on Implicit Function Representation of Light Fields Using Deep Learning
基于深度学习光场隐式函数表示的头戴式显示器空间校准
- 批准号:
23K16920 - 财政年份:2023
- 资助金额:
$ 29.5万 - 项目类别:
Grant-in-Aid for Early-Career Scientists
Peer suppoRt for adolescents and Emerging adults with Sickle cell pain: promoting ENgagement in Cognitive behavioral thErapy
对患有镰状细胞痛的青少年和新兴成年人的同伴支持:促进认知行为治疗的参与
- 批准号:
10745058 - 财政年份:2023
- 资助金额:
$ 29.5万 - 项目类别:
IC-CSRisk Study: Implementation of Calculated Cesarean Section Risk during labor induction, a multi-site stepped-wedge randomized rollout trial
IC-CSRisk 研究:引产期间计算剖腹产风险的实施,一项多站点阶梯式楔形随机推广试验
- 批准号:
10776163 - 财政年份:2023
- 资助金额:
$ 29.5万 - 项目类别:
African American (AA) Communities Speak: Partnering with AAs in the North and South to Train Palliative Care Clinicians to Address Interpersonal and Systemic Racism and Provide Culturally Aligned Care
非裔美国人 (AA) 社区发言:与北部和南部的 AA 合作,培训姑息治疗临床医生,以解决人际和系统性种族主义并提供文化一致的护理
- 批准号:
10734272 - 财政年份:2023
- 资助金额:
$ 29.5万 - 项目类别: