III: Small: Declarative Recursive Computation on a Database System

III：小型：数据库系统上的声明式递归计算

基本信息

批准号：
1910803
负责人：
Christopher Jermaine
金额：
$ 50万
依托单位：
William Marsh Rice University
依托单位国家：
美国
项目类别：
Standard Grant
财政年份：
2019
资助国家：
美国
起止时间：
2019-08-15 至 2023-07-31
项目状态：
已结题

来源：
https://www.nsf.gov/awardsearch/showAward?AWD_ID=1910803&HistoricalAwards=false
关键词：
III Small Declarative Recursive Computation

项目摘要

Machine learning (ML) will have a huge economic and scientific impact over the upcoming decades. But despite the growing importance of ML, there has been relatively little work focused on asking what systems to support machine learning should look like. The result is that it is difficult to apply ML to non-standard cases: Big Data, large or complex models that take a long time to train or require more RAM than is available on a graphics processing unit (GPU), or learning problems with hard training time constraints, to name a few. The proposed project aims to address such deficiencies by applying ideas from relational database systems to the design and implementation of systems for ML. Building an ML system on top of a relational-style engine will enable the design of ML systems that are able to automatically generate compute plans for specific ML tasks with little programmer effort. Those plans will be optimized and executed to match the data size, layout, and the compute hardware. The code to implement an ML algorithm will be the same no matter whether the computation is run on a local machine, or in a distributed environment. If successful, the project will radically expand the ease-of-use and applicability of ML.There are a number of technical questions that need to be answered for ML computations to be run on top of a relational system, and answering such questions will be at the heart of the project. For example: How can ML primitives (convolutions, recurrent modules, etc.) be mapped onto relational primitives? How can large objects (matrices/tensors) be chucked into records so relational implementations run efficiently? And since a developer of ML algorithms is unlikely to accept SQL as a programming language: How to translate Karas-like Python programs into relational algebra?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.

机器学习 (ML) 将在未来几十年产生巨大的经济和科学影响。但是，尽管机器学习的重要性与日俱增，但关注支持机器学习的系统应该是什么样子的工作却相对较少。结果是，很难将 ML 应用于非标准情况：大数据、大型或复杂模型需要很长时间来训练或需要比图形处理单元 (GPU) 上可用的更多 RAM，或者学习问题艰苦的训练时间限制等等。拟议的项目旨在通过将关系数据库系统的思想应用于机器学习系统的设计和实现来解决这些缺陷。在关系型引擎之上构建 ML 系统将使 ML 系统的设计能够自动为特定 ML 任务生成计算计划，而程序员只需付出很少的努力。这些计划将被优化和执行，以匹配数据大小、布局和计算硬件。无论计算是在本地计算机上还是在分布式环境中运行，实现机器学习算法的代码都是相同的。如果成功，该项目将从根本上扩展机器学习的易用性和适用性。要在关系系统之上运行机器学习计算，需要回答许多技术问题，而回答这些问题将是是该项目的核心。例如：如何将 ML 原语（卷积、循环模块等）映射到关系原语？如何将大型对象（矩阵/张量）放入记录中以便关系实现高效运行？而且，由于 ML 算法的开发人员不太可能接受 SQL 作为编程语言：如何将类似 Karas 的 Python 程序转化为关系代数？该奖项反映了 NSF 的法定使命，并通过使用基金会的智力优点和能力进行评估，被认为值得支持。更广泛的影响审查标准。

项目成果

期刊论文数量（6）

专著数量（0）

科研奖励数量（0）

会议论文数量（0）

专利数量（0）

Distributed Learning of Fully Connected Neural Networks using Independent Subnet Training

DOI：
10.14778/3529337.3529343
发表时间：
2019-10
期刊：
Proc. VLDB Endow.
影响因子：
0
作者：
Binhang Yuan;Anastasios Kyrillidis;C. Jermaine
通讯作者：
Binhang Yuan;Anastasios Kyrillidis;C. Jermaine

Tensor Relational Algebra for Distributed Machine Learning System Design

DOI：
10.14778/3457390.3457399
发表时间：
2020-09
期刊：
Proc. VLDB Endow.
影响因子：
0
作者：
Binhang Yuan;Dimitrije Jankov;Jia Zou;Yu-Shuen Tang;Daniel Bourgeois;C. Jermaine
通讯作者：
Binhang Yuan;Dimitrije Jankov;Jia Zou;Yu-Shuen Tang;Daniel Bourgeois;C. Jermaine

MONSOON: Multi-Step Optimization and Execution of Queries with Partially Obscured Predicates

MONSOON：使用部分模糊谓词的查询的多步优化和执行