FMitF:Track I: Verified Safe and Fair Machine Learning

FMITF:第一轨:经过验证的安全和公平的机器学习

基本信息

  • 批准号:
    2018372
  • 负责人:
  • 金额:
    $ 74.99万
  • 依托单位:
  • 依托单位国家:
    美国
  • 项目类别:
    Standard Grant
  • 财政年份:
    2020
  • 资助国家:
    美国
  • 起止时间:
    2020-10-01 至 2024-09-30
  • 项目状态:
    已结题

项目摘要

Artificial intelligence (AI), and specifically machine learning, is being used more and more in areas with significant real-world impacts on people's lives. Examples include the delivery of health care and social services, decisions in the legal-justice system, self-driving cars, and face and speech recognition. Researchers have discovered that these applications of machine learning often embody biases, or health, safety, or economic risks. This project's novelty lies in developing ways to show that a test of the safety or fairness of a machine-learning system is mathematically sound and correctly coded on a computer, so that its test results can be relied upon. The project's impacts will thus be greater assurance that risks (lack of safety) and biases (lack of fairness) are known and evaluated precisely and correctly.The investigators develop computer-checked proofs of correctness of several components necessary to the overall goals described above. These computer-checked proofs of formulations of the necessary statistical tests, such as Hoeffding's Inequality (and other such inequalities), are used to bound the probability that bias or safety risk exceeds a given limit. The mathematics of these is known, but computer-checked proofs are novel. Further, some newer bounds have hand-written proofs possibly needing more rigor or stronger assumptions, the limitations of which will be revealed by attempting computer-checked proofs. Next, computer code used to implement the safety/fairness tests needs similar proofs of correctness. Some aspects of how to do this are well-known, but computer-checked proofs for the numerical (floating-point) computations involved are lacking, and challenging. Lastly, the researchers will improve the computer proof tools, which remain weak in certain respects, by using machine learning to assist in these kinds of proofs.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.
人工智能(AI),特别是机器学习,正越来越多地用于对人们生活产生重大现实影响的领域。 例如,医疗保健和社会服务的提供,法律司法系统的决策,自动驾驶汽车以及面部和语音识别。 研究人员发现,机器学习的这些应用通常包含偏见,或健康,安全或经济风险。 该项目的新奇在于开发方法来证明机器学习系统的安全性或公平性测试在数学上是合理的,并且在计算机上正确编码,因此其测试结果可以信赖。 因此,项目的影响将是更大的保证,风险(缺乏安全性)和偏见(缺乏公平性)是已知的,并准确和正确的评估。调查人员开发的计算机检查的几个组件的正确性的证据,以实现上述的总体目标。 这些必要的统计检验公式的计算机检查证明,如Hoeffding不等式(和其他此类不等式),用于限制偏倚或安全风险超过给定限值的概率。 这些数学是已知的,但计算机检查的证明是新颖的。 此外,一些较新的界限有可能需要更严格或更强的假设的手写证明,其局限性将通过尝试计算机检查证明来揭示。接下来,用于实现安全/公平性测试的计算机代码需要类似的正确性证明。如何做到这一点的一些方面是众所周知的,但所涉及的数值(浮点)计算的计算机检查证明是缺乏的,并且具有挑战性。 最后,研究人员将通过使用机器学习来帮助进行此类证明,从而改进在某些方面仍然薄弱的计算机证明工具。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

期刊论文数量(9)
专著数量(0)
科研奖励数量(0)
会议论文数量(0)
专利数量(0)
Fairness Guarantees under Demographic Shift
Security Analysis of Safe and Seldonian Reinforcement Learning Algorithms
安全和Seldonian强化学习算法的安全性分析
Universal Off-Policy Evaluation
  • DOI:
  • 发表时间:
    2021-04
  • 期刊:
  • 影响因子:
    0
  • 作者:
    Yash Chandak;S. Niekum;Bruno C. da Silva;E. Learned-Miller;E. Brunskill;P. Thomas
  • 通讯作者:
    Yash Chandak;S. Niekum;Bruno C. da Silva;E. Learned-Miller;E. Brunskill;P. Thomas
Towards Practical Mean Bounds for Small Samples
走向小样本的实际平均界限
Mechanizing Soundness of Off-Policy Evaluation
  • DOI:
    10.4230/lipics.itp.2022.32
  • 发表时间:
    2022
  • 期刊:
  • 影响因子:
    0
  • 作者:
    Jared Yeager;J. Moss;Michael Norrish;P. Thomas
  • 通讯作者:
    Jared Yeager;J. Moss;Michael Norrish;P. Thomas
{{ 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 }}

J. Eliot Moss其他文献

J. Eliot Moss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('J. Eliot Moss', 18)}}的其他基金

CNS Core: Small: Managed Languages: From Non-volatile Memory to Persistence
CNS 核心:小型:托管语言:从非易失性内存到持久性
  • 批准号:
    1909731
  • 财政年份:
    2019
  • 资助金额:
    $ 74.99万
  • 项目类别:
    Standard Grant
SHF: Medium: Collaborative Research: Micro-Virtual Machines for Managed Languages: Abstraction, contained
SHF:媒介:协作研究:托管语言的微型虚拟机:抽象,包含
  • 批准号:
    1832624
  • 财政年份:
    2017
  • 资助金额:
    $ 74.99万
  • 项目类别:
    Standard Grant
CSR: Medium: Collaborative Research: Portable Performance for Parallel Managed Languages Across the Many-Core Spectrum
CSR:媒介:协作研究:跨多核频谱的并行托管语言的可移植性能
  • 批准号:
    1833291
  • 财政年份:
    2017
  • 资助金额:
    $ 74.99万
  • 项目类别:
    Continuing Grant
SHF: Medium: Collaborative Research: Micro Virtual Machines for Managed Languages: Abstraction, defined and contained
SHF:媒介:协作研究:托管语言的微型虚拟机:抽象、定义和包含
  • 批准号:
    1409284
  • 财政年份:
    2014
  • 资助金额:
    $ 74.99万
  • 项目类别:
    Standard Grant
SHF:Small: Accurate and Computationally Efficient Predictors of Java Memory Resource Consumption
SHF:Small:Java 内存资源消耗的准确且计算高效的预测器
  • 批准号:
    1320498
  • 财政年份:
    2013
  • 资助金额:
    $ 74.99万
  • 项目类别:
    Standard Grant
CSR: Medium: Collaborative Research: Portable Performance for Parallel Managed Languages Across the Many-Core Spectrum
CSR:媒介:协作研究:跨多核频谱的并行托管语言的可移植性能
  • 批准号:
    1162246
  • 财政年份:
    2012
  • 资助金额:
    $ 74.99万
  • 项目类别:
    Continuing Grant
EAGER: Automating Correctness Proofs of Transactionalized Data Structures
EAGER:自动化事务化数据结构的正确性证明
  • 批准号:
    0953761
  • 财政年份:
    2009
  • 资助金额:
    $ 74.99万
  • 项目类别:
    Standard Grant
Describing the Operating System for Accurate User-mode Simulation
描述用于精确用户模式模拟的操作系统
  • 批准号:
    0950410
  • 财政年份:
    2009
  • 资助金额:
    $ 74.99万
  • 项目类别:
    Standard Grant
SGER: The Chaotic Behavior of Automatic Memory Management
SGER:自动内存管理的混乱行为
  • 批准号:
    0836542
  • 财政年份:
    2008
  • 资助金额:
    $ 74.99万
  • 项目类别:
    Standard Grant
CSR-AES Collaborative: Encore/J: Transparently Recoverable Java for Resilient Distributed Computing
CSR-AES 协作:Encore/J:用于弹性分布式计算的透明可恢复 Java
  • 批准号:
    0720242
  • 财政年份:
    2007
  • 资助金额:
    $ 74.99万
  • 项目类别:
    Standard Grant

相似海外基金

RII Track-4:NSF: Integrated Electrochemical-Optical Microscopy for High Throughput Screening of Electrocatalysts
RII Track-4:NSF:用于高通量筛选电催化剂的集成电化学光学显微镜
  • 批准号:
    2327025
  • 财政年份:
    2024
  • 资助金额:
    $ 74.99万
  • 项目类别:
    Standard Grant
RII Track-4:NSF: Resistively-Detected Electron Spin Resonance in Multilayer Graphene
RII Track-4:NSF:多层石墨烯中电阻检测的电子自旋共振
  • 批准号:
    2327206
  • 财政年份:
    2024
  • 资助金额:
    $ 74.99万
  • 项目类别:
    Standard Grant
RII Track-4:NSF: Improving subseasonal-to-seasonal forecasts of Central Pacific extreme hydrometeorological events and their impacts in Hawaii
RII Track-4:NSF:改进中太平洋极端水文气象事件的次季节到季节预报及其对夏威夷的影响
  • 批准号:
    2327232
  • 财政年份:
    2024
  • 资助金额:
    $ 74.99万
  • 项目类别:
    Standard Grant
RII Track-4:NSF: Design of zeolite-encapsulated metal phthalocyanines catalysts enabled by insights from synchrotron-based X-ray techniques
RII Track-4:NSF:通过基于同步加速器的 X 射线技术的见解实现沸石封装金属酞菁催化剂的设计
  • 批准号:
    2327267
  • 财政年份:
    2024
  • 资助金额:
    $ 74.99万
  • 项目类别:
    Standard Grant
RII Track-4:NSF: From the Ground Up to the Air Above Coastal Dunes: How Groundwater and Evaporation Affect the Mechanism of Wind Erosion
RII Track-4:NSF:从地面到沿海沙丘上方的空气:地下水和蒸发如何影响风蚀机制
  • 批准号:
    2327346
  • 财政年份:
    2024
  • 资助金额:
    $ 74.99万
  • 项目类别:
    Standard Grant
RII Track-4:NSF: In-Situ/Operando Characterizations of Single Atom Catalysts for Clean Fuel Generation
RII Track-4:NSF:用于清洁燃料生成的单原子催化剂的原位/操作表征
  • 批准号:
    2327349
  • 财政年份:
    2024
  • 资助金额:
    $ 74.99万
  • 项目类别:
    Standard Grant
RII Track-4: NSF: Fundamental study on hydrogen flow in porous media during repetitive drainage-imbibition processes and upscaling for underground energy storage
RII Track-4:NSF:重复排水-自吸过程中多孔介质中氢气流动的基础研究以及地下储能的升级
  • 批准号:
    2327317
  • 财政年份:
    2024
  • 资助金额:
    $ 74.99万
  • 项目类别:
    Standard Grant
RII Track-4:@NASA: Wind-induced noise in the prospective seismic data measured in the Venusian surface environment
RII Track-4:@NASA:金星表面环境中测量的预期地震数据中的风致噪声
  • 批准号:
    2327422
  • 财政年份:
    2024
  • 资助金额:
    $ 74.99万
  • 项目类别:
    Standard Grant
RII Track-4:NSF: An Integrated Urban Meteorological and Building Stock Modeling Framework to Enhance City-level Building Energy Use Predictions
RII Track-4:NSF:综合城市气象和建筑群建模框架,以增强城市级建筑能源使用预测
  • 批准号:
    2327435
  • 财政年份:
    2024
  • 资助金额:
    $ 74.99万
  • 项目类别:
    Standard Grant
RII Track-4: NSF: Developing 3D Models of Live-Endothelial Cell Dynamics with Application Appropriate Validation
RII Track-4:NSF:开发活内皮细胞动力学的 3D 模型并进行适当的应用验证
  • 批准号:
    2327466
  • 财政年份:
    2024
  • 资助金额:
    $ 74.99万
  • 项目类别:
    Standard Grant
{{ showInfoDetail.title }}

作者:{{ showInfoDetail.author }}

知道了