CAREER: Scaling Approximate Inference and Approximation-Aware Learning

职业：扩展近似推理和近似感知学习

• 批准号: 1762268
• 负责人: Wolfgang Gatterbauer
• 金额: $ 53.14万
• 依托单位: Northeastern University
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2017
• 资助国家: 美国
• 起止时间: 2017-09-01 至 2025-03-31
• 项目状态: 未结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1762268&HistoricalAwards=false
• 关键词: CAREER; Scaling; Approximate; Inference; Approximation

The last decade has seen an enormous increase in our ability to gather and manage large amounts of data; business, healthcare, education, economy, science, and almost every aspect of society are accumulating data at unprecedented levels. The basic premise is that by having more data, even if uncertain and of lower quality, we are also able to make better-informed decisions. To make any decisions, we need to perform "inference" over the data, i.e. to either draw new conclusions, or to find support for existing hypotheses, thus allowing us to favor one course of action over another. However, general reasoning under uncertainty is highly intractable, and many state-of-the-art systems today perform approximate inference by reverting to sampling. Thus for many modern applications (such as information extraction, knowledge aggregation, question-answering systems, computer vision, and machine intelligence), inference is a key bottleneck, and new methods for tractable approximate inference are needed.This project addresses the challenge of scaling inference by generalizing two highly scalable approximate inference methods and complementing them with scalable methods for parameter learning that are "approximation-aware." Thus, instead of treating the (i) learning and the (ii) inference steps separately, this project uses the approximation methods developed for inference also for learning the model. The research hypothesis is that this approach increases the overall end-to-end prediction accuracy while simultaneously increasing scalability. Concretely, the project develops the theory and a set of scalable algorithms and optimization methods for at least the following four sub-problems: (1) approximating general probabilistic conjunctive queries with standard relational databases; (2) learning the probabilities in uncertain databases based on feedback on rankings of output tuples from general queries; (3) approximating the exact probabilistic inference in undirected graphical models with linearized update equations; and (4) complementing the latter with a robust framework for learning linearized potentials from partially labeled data.

在过去的十年里，我们收集和管理大量数据的能力有了巨大的提高；商业、医疗、教育、经济、科学以及社会的几乎每个方面都在以前所未有的水平积累数据。基本的前提是，通过拥有更多的数据，即使不确定和质量较低，我们也能够做出更明智的决定。为了做出任何决定，我们需要对数据进行“推断”，即要么得出新的结论，要么为现有的假设找到支持，从而使我们能够倾向于一种行动方案而不是另一种。然而，不确定性下的一般推理是非常棘手的，今天许多最先进的系统通过恢复采样来进行近似推理。因此，对于许多现代应用（如信息提取、知识聚合、问答系统、计算机视觉和机器智能），推理是一个关键的瓶颈，需要新的易于处理的近似推理方法。该项目通过推广两种高度可扩展的近似推理方法，并使用“近似感知”的参数学习可扩展方法来补充它们，解决了缩放推理的挑战。因此，本项目不是单独处理(i)学习和（ii）推理步骤，而是使用为推理开发的近似方法来学习模型。研究假设是，这种方法提高了整体的端到端预测精度，同时提高了可扩展性。具体而言，该项目为至少以下四个子问题开发了理论和一套可扩展的算法和优化方法：(1)用标准关系数据库近似一般概率联合查询；(2)基于对一般查询输出元组排序的反馈，学习不确定数据库中的概率；(3)用线性化的更新方程逼近无向图模型的精确概率推理；(4)用一个鲁棒框架从部分标记数据中学习线性化电位来补充后者。

项目成果

Tractable Orders for Direct Access to Ranked Answers of Conjunctive Queries

用于直接访问连接查询的排名答案的易于处理的顺序

• DOI: 10.1145/3578517
• 发表时间: 2023
• 期刊: ACM Transactions on Database Systems
• 影响因子: 1.8
• 作者: Carmeli, Nofar;Tziavelis, Nikolaos;Gatterbauer, Wolfgang;Kimelfeld, Benny;Riedewald, Mirek
• 通讯作者: Riedewald, Mirek

DomainNet: Homograph Detection for Data Lake Disambiguation

• DOI: 10.5441/002/edbt.2021.03
• 发表时间: 2021-03
• 期刊: ArXiv
• 作者: Aristotelis Leventidis;Laura Di Rocco;Wolfgang Gatterbauer;Renée J. Miller;Mirek Riedewald

Efficient Computation of Quantiles over Joins

通过连接高效计算分位数

• DOI: 10.1145/3584372.3588670
• 发表时间: 2023
• 期刊: PODS
• 作者: Tziavelis, Nikolaos;Carmeli, Nofar;Gatterbauer, Wolfgang;Kimelfeld, Benny;Riedewald, Mirek
• 通讯作者: Riedewald, Mirek

Toward Responsive DBMS: Optimal Join Algorithms, Enumeration, Factorization, Ranking, and Dynamic Programming

迈向响应式 DBMS：最优连接算法、枚举、因式分解、排序和动态规划

• DOI: 10.1109/icde53745.2022.00299
• 发表时间: 2022
• 期刊: ICDE tutorials
• 作者: Tziavelis, Nikolaos;Gatterbauer, Wolfgang;Riedewald, Mirek
• 通讯作者: Riedewald, Mirek

SANTOS: Relationship-based Semantic Table Union Search

SANTOS：基于关系的语义表联合搜索

• DOI: 10.1145/3588689
• 发表时间: 2023
• 期刊: Proceedings of the ACM on Management of Data
• 作者: Khatiwada, Aamod;Fan, Grace;Shraga, Roee;Chen, Zixuan;Gatterbauer, Wolfgang;Miller, Renée J.;Riedewald, Mirek
• 通讯作者: Riedewald, Mirek