FMitF: Opening Up the Black Box of Probabilistic Program Inference
FMITF:打开概率程序推理的黑匣子
基本信息
- 批准号:1837129
- 负责人:
- 金额:$ 94.74万
- 依托单位:
- 依托单位国家:美国
- 项目类别:Standard Grant
- 财政年份:2018
- 资助国家:美国
- 起止时间:2018-12-01 至 2023-11-30
- 项目状态:已结题
- 来源:
- 关键词:
项目摘要
Probabilistic programming languages are an expressive means for creating, maintaining, and understanding a wide range of machine-learning models, and they have been successfully used by both researchers and major technology companies. However, today's probabilistic programming languages impose strong limitations on the kinds of programs for which they are effective, thereby precluding their use for many machine-learning applications. This project adapts and generalizes techniques from the formal methods community for reasoning about traditional programs, in order to develop general-purpose algorithms for probabilistic inference, which is the key task that a probabilistic programming language must perform. The project is implementing these algorithms in the context of a new imperative probabilistic programming language and is providing educational opportunities in the burgeoning area of probabilistic programming for graduate, undergraduate, and high-school students.This project has three main technical thrusts. First, the project is developing exact inference algorithms for discrete probabilistic programs by exploiting the connection to techniques for symbolic model checking and weighted model counting. Second, the project is developing techniques to automatically decompose a probabilistic inference query into multiple simpler sub-queries, each of which can be solved using the most appropriate inference method. The key enabler of this decomposition is a novel abstraction process for probabilistic programs. Third, the project is investigating the use of abstractions as proposal distributions for probabilistic inference, resulting in new abstraction-guided approximate inference algorithms. The results of this project will make probabilistic programming more effective by making probabilistic inference and learning tractable for a wider class of programs. The artifacts that result from this research will be released open source, including a new probabilistic programming language that leverages the newly developed inference techniques.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.
概率编程语言是一种用于创建、维护和理解各种机器学习模型的表达手段,它们已被研究人员和主要技术公司成功使用。 然而,今天的概率编程语言对它们有效的程序类型施加了很大的限制,从而排除了它们在许多机器学习应用中的使用。 该项目适应和推广的技术,从形式化方法社区推理传统的程序,以开发通用的算法概率推理,这是一个概率编程语言必须执行的关键任务。 该项目正在实现这些算法的背景下,一个新的命令式概率编程语言,并提供教育机会,在新兴领域的概率编程的研究生,本科生和高中生。 首先,该项目是开发精确的推理算法的离散概率程序,利用连接到技术的符号模型检查和加权模型计数。 第二,该项目正在开发技术,以自动将概率推理查询分解为多个更简单的子查询,每个子查询都可以使用最合适的推理方法来解决。这种分解的关键使能器是概率程序的一种新的抽象过程。 第三,该项目正在研究使用抽象作为概率推理的建议分布,从而产生新的抽象指导的近似推理算法。 这个项目的结果将使概率编程更有效的概率推理和学习更广泛的程序类听话。 这项研究的成果将开源发布,包括一种新的概率编程语言,该语言利用了新开发的推理技术。该奖项反映了NSF的法定使命,并通过使用基金会的知识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(13)
专著数量(0)
科研奖励数量(0)
会议论文数量(0)
专利数量(0)
Data-driven inference of representation invariants
表示不变量的数据驱动推理
- DOI:10.1145/3385412.3385967
- 发表时间:2020
- 期刊:
- 影响因子:0
- 作者:Miltner, Anders;Padhi, Saswat;Millstein, Todd;Walker, David
- 通讯作者:Walker, David
Counterexample-Guided Learning of Monotonic Neural Networks
- DOI:
- 发表时间:2020-06
- 期刊:
- 影响因子:0
- 作者:Aishwarya Sivaraman;G. Farnadi;T. Millstein;Guy Van den Broeck
- 通讯作者:Aishwarya Sivaraman;G. Farnadi;T. Millstein;Guy Van den Broeck
Scaling Integer Arithmetic in Probabilistic Programs
概率程序中整数算术的缩放
- DOI:
- 发表时间:2023
- 期刊:
- 影响因子:0
- 作者:Cao, William;Garg, Poorva;Tjoa, Ryan;Holtzen, Steven;Millstein, Todd;Van den Broeck, Guy
- 通讯作者:Van den Broeck, Guy
Data-driven lemma synthesis for interactive proofs
用于交互式证明的数据驱动引理合成
- DOI:10.1145/3563306
- 发表时间:2022
- 期刊:
- 影响因子:0
- 作者:Sivaraman, Aishwarya;Sanchez-Stern, Alex;Chen, Bretton;Lerner, Sorin;Millstein, Todd
- 通讯作者:Millstein, Todd
Efficient Search-Based Weighted Model Integration
基于搜索的高效加权模型集成
- DOI:
- 发表时间:2019
- 期刊:
- 影响因子:0
- 作者:Zeng, Z.;Van den Broeck, G.
- 通讯作者:Van den Broeck, G.
{{
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 }}
Todd Millstein其他文献
Todd Millstein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('Todd Millstein', 18)}}的其他基金
Collaborative Research: SHF: Small: Data-Driven Lemma Synthesis for Interactive Proofs
协作研究:SHF:小型:交互式证明的数据驱动引理合成
- 批准号:
2220891 - 财政年份:2022
- 资助金额:
$ 94.74万 - 项目类别:
Standard Grant
QCIS-FF: A Software Stack for Quantum Computing
QCIS-FF:量子计算软件堆栈
- 批准号:
1926648 - 财政年份:2020
- 资助金额:
$ 94.74万 - 项目类别:
Continuing Grant
NeTS: Medium: Collaborative Research: Network Configuration Synthesis: A Path to Practical Deployment
NeTS:媒介:协作研究:网络配置综合:实际部署之路
- 批准号:
1704336 - 财政年份:2017
- 资助金额:
$ 94.74万 - 项目类别:
Continuing Grant
SHF: Small: Interacting to Specify Software
SHF:小型:交互指定软件
- 批准号:
1527923 - 财政年份:2015
- 资助金额:
$ 94.74万 - 项目类别:
Standard Grant
NeTS: Medium: Collaborative Research: Systematic Analysis of Protocol Implementations
NeTS:媒介:协作研究:协议实现的系统分析
- 批准号:
1161595 - 财政年份:2012
- 资助金额:
$ 94.74万 - 项目类别:
Continuing Grant
TC: Medium: Collaborative Research: Program Analysis for Smartphone Application Security
TC:媒介:协作研究:智能手机应用程序安全的程序分析
- 批准号:
1064844 - 财政年份:2011
- 资助金额:
$ 94.74万 - 项目类别:
Standard Grant
EAGER: Collaborative Research: Toward An Adaptive Programming System for Cloud-Enabled Smartphone Applications
EAGER:协作研究:面向云智能手机应用程序的自适应编程系统
- 批准号:
1048826 - 财政年份:2010
- 资助金额:
$ 94.74万 - 项目类别:
Standard Grant
SoD: An Electronic Design Automation Approach to Embedded Networked Software
SoD:嵌入式网络软件的电子设计自动化方法
- 批准号:
0725354 - 财政年份:2007
- 资助金额:
$ 94.74万 - 项目类别:
Standard Grant
"CAREER:" Enforcing and Validating User-Defined Programming Disciplines
“职业:”执行和验证用户定义的编程规则
- 批准号:
0545850 - 财政年份:2006
- 资助金额:
$ 94.74万 - 项目类别:
Continuing Grant
相似国自然基金
萱草花开放时间(Flower Opening Time)的生物钟调控机制研究
- 批准号:31971706
- 批准年份:2019
- 资助金额:59.0 万元
- 项目类别:面上项目
相似海外基金
Opening up Anopheles funestus to functional genetics and the study of insecticide resistance
开启按蚊的功能遗传学和杀虫剂抗性研究
- 批准号:
MR/Y002008/1 - 财政年份:2024
- 资助金额:
$ 94.74万 - 项目类别:
Research Grant
Make Film History: Opening up the BBC Archives to Young Filmmakers
创造电影历史:向年轻电影制作人开放 BBC 档案
- 批准号:
AH/X001156/1 - 财政年份:2022
- 资助金额:
$ 94.74万 - 项目类别:
Research Grant
Opening Up Access to L-Sugars through a Synergy of Experiment and Theory
通过实验和理论的协同作用开放 L-糖的获取
- 批准号:
DP220102493 - 财政年份:2022
- 资助金额:
$ 94.74万 - 项目类别:
Discovery Projects
How plants open up: revealing the evolution of stomatal opening mechanisms
植物如何打开:揭示气孔打开机制的演变
- 批准号:
DP220101795 - 财政年份:2022
- 资助金额:
$ 94.74万 - 项目类别:
Discovery Projects
Opening Up Minds: Engaging Dialogue Generated From Argument Maps
开放思想:通过论证图进行对话
- 批准号:
EP/T023414/1 - 财政年份:2021
- 资助金额:
$ 94.74万 - 项目类别:
Research Grant
Opening Up Minds: Engaging Dialogue Generated From Argument Maps
开放思想:通过论证图进行对话
- 批准号:
EP/T024666/1 - 财政年份:2021
- 资助金额:
$ 94.74万 - 项目类别:
Research Grant
Opening Up Minds: Engaging Dialogue Generated From Argument Maps
开放思想:通过论证图进行对话
- 批准号:
EP/T023554/1 - 财政年份:2021
- 资助金额:
$ 94.74万 - 项目类别:
Research Grant
Make Film History - Opening up the Archives to Young Filmmakers
创造电影历史——向年轻电影制片人开放档案
- 批准号:
AH/V002066/1 - 财政年份:2020
- 资助金额:
$ 94.74万 - 项目类别:
Research Grant
Opening up New Oxidation States in Metallocene Chemistry Across the Periodic Table
在整个元素周期表的茂金属化学中开辟新的氧化态
- 批准号:
2297278 - 财政年份:2019
- 资助金额:
$ 94.74万 - 项目类别:
Studentship
Case study of Chinese governments' opening-up policies in higher education
中国政府高等教育开放政策案例研究
- 批准号:
18K11776 - 财政年份:2018
- 资助金额:
$ 94.74万 - 项目类别:
Grant-in-Aid for Scientific Research (C)