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：
J. Eliot Moss
金额：
$ 74.99万
依托单位：
University of Massachusetts Amherst
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

来源：
https://www.nsf.gov/awardsearch/showAward?AWD_ID=2018372&HistoricalAwards=false
关键词：
FMitF Track Verified Safe Fair

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，越来越多地应用于对人们的生活产生重大现实影响的领域。例子包括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存在偏见，或者健康、安全或经济风险。该项目的新颖之处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，该项目的影响将是更好地确保风险（缺乏安全性）和偏见（缺乏公平性）得到了解并进行精确和正确的评估。研究人员开发了计算机检查的证明，以证明实现上述总体目标所必需的几个组件的正确性。这些经过计算机检查的必要统计测试公式的证明，例如霍夫丁不等式（以及其他此类不等式），用于限制偏差或安全风险超过给定限制的概率。这些的数学原理是已知的，但计算机检查的证明是新颖的。此外，一些较新的边界具有手写证明，可能需要更严格或更强有力的假设，其局限性将通过尝试计算机检查的证明来揭示。接下来，用于实现安全/公平测试的计算机代码需要类似的正确性证明。如何做到这一点的某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却缺乏且具有挑战性。最后，研究人员将通过使用机器学习来辅助此类证明，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进行评估，认为值得支持以及更广泛的影响审查标准。

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

DOI：
10.4108/eai.12-1-2024.2347147
发表时间：
2022
期刊：
Proceedings of the 3rd International Conference on Big Data Economy and Digital Management, BDEDM 2024, January 12–14, 2024, Ningbo, China
影响因子：
0
作者：
Stephen Giguere;Blossom Metevier;B. C. Silva;Yuriy Brun;P. Thomas;S. Niekum
通讯作者：
Stephen Giguere;Blossom Metevier;B. C. Silva;Yuriy Brun;P. Thomas;S. Niekum

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

DOI：
发表时间：
2020
期刊：
Advances in neural information processing systems
影响因子：
0
作者：
Ozisik, Pinar;Thomas, Philip
通讯作者：
Thomas, Philip

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