SHF: Small: Collaborative Research: Programming Tools for Adaptive Data Analysis

SHF：小型：协作研究：自适应数据分析的编程工具

基本信息

批准号：
1718220
负责人：
Marco Gaboardi
金额：
$ 22.45万
依托单位：
SUNY at Buffalo
依托单位国家：
美国
项目类别：
Standard Grant
财政年份：
2017
资助国家：
美国
起止时间：
2017-08-01 至 2020-09-30
项目状态：
已结题

来源：
https://www.nsf.gov/awardsearch/showAward?AWD_ID=1718220&HistoricalAwards=false
关键词：
SHF Small Collaborative Research Programming

项目摘要

False discovery, or overfitting, occurs when an empirical researcher draws a conclusion based on a dataset that does not generalize to new data. Although there are many statistical methods for preventing false discovery, most are designed for static data analysis, where a dataset is used only once. However, modern data analysis is adaptive, and often the same datasets are reused for multiple studies by multiple researchers. Adaptivity has been identified by statisticians as one cause of non-reproducible research, and this project?s broader significance and importance will be to begin addressing this problem. Specifically, this project will build a prototype programming tool for preventing false discovery arising from adaptive data analysis. The intellectual merits are to incorporate and extend recent theoretical advances on this problem into a programming framework that allows researchers to analyze datasets adaptively with robust guarantees that overfitting will not occur.The project builds on a surprising recent connection between differential privacy and false discovery, a robust statistical guarantee that emerged recently to protect the privacy of sensitive data. This line of work shows that when data is analyzed in a differentially private way, then false discoveries cannot occur. Differential privacy is also programmable, and allows complex differentially private algorithms to be built from simple components, so it is an ideal programming framework for adaptive data analysis. This project is extending existing differentially private programming frameworks to adaptive data analysis. The PIs are also developing new algorithmic and programming languages tools for adaptive data analysis, and incorporating them into the first prototype system for this application.

当经验研究人员根据数据集提出一个不会推广到新数据的数据集的结论时，会发生错误的发现或过度拟合。尽管有许多用于防止错误发现的统计方法，但大多数是为静态数据分析而设计的，其中仅使用一次数据集。但是，现代数据分析是自适应的，并且通常由多个研究人员重复使用相同的数据集进行多项研究。统计学家已经将适应性确定为不可复制研究的原因之一，该项目的重要性和重要性将是开始解决这个问题。具体而言，该项目将构建一个原型编程工具，以防止自适应数据分析引起的错误发现。智力优点是将此问题上的最新理论进步纳入一个编程框架中，该框架使研究人员可以通过强大的保证可以适应数据集，以确保不会发生过度拟合。这项工作表明，当以差异性私有方式分析数据时，就不会发生错误的发现。差异隐私也可以编程，并允许通过简单组件构建复杂的差异化私有算法，因此它是自适应数据分析的理想编程框架。该项目正在将现有的不同私有编程框架扩展到适应性数据分析。 PI还为自适应数据分析开发了新的算法和编程语言工具，并将其纳入该应用程序的第一个原型系统。

项目成果

期刊论文数量（14）

专著数量（0）

科研奖励数量（0）

会议论文数量（0）

专利数量（0）

Probabilistic Relational Reasoning via Metrics

DOI：
10.1109/lics.2019.8785715
发表时间：
2019-01-01
期刊：
2019 34TH ANNUAL ACM/IEEE SYMPOSIUM ON LOGIC IN COMPUTER SCIENCE (LICS)
影响因子：
0
作者：
de Amorim, Arthur Azevedo;Gaboardi, Marco;Katsumata, Shin-ya
通讯作者：
Katsumata, Shin-ya

Relational cost analysis for functional-imperative programs

功能命令式程序的关系成本分析

DOI：
10.1145/3341696
发表时间：
2019
期刊：
Proceedings of the ACM on programming languages
影响因子：
0
作者：
Qu, Weihao;Gaboardi, Marco;Garg, Deepak
通讯作者：
Garg, Deepak

Facility Location Problem in Differential Privacy Model Revisited

DOI：
发表时间：
2019-10
期刊：
影响因子：
0
作者：
Yunus Esencayi;Marco Gaboardi;Shi Li;Di Wang
通讯作者：
Yunus Esencayi;Marco Gaboardi;Shi Li;Di Wang

Covariance-Aware Private Mean Estimation Without Private Covariance Estimation

DOI：
发表时间：
2021-06
期刊：
ArXiv
影响因子：
0
作者：
Gavin Brown;Marco Gaboardi;Adam D. Smith;Jonathan Ullman;Lydia Zakynthinou
通讯作者：
Gavin Brown;Marco Gaboardi;Adam D. Smith;Jonathan Ullman;Lydia Zakynthinou

A Programming Language for Data Privacy with Accuracy Estimations

具有准确性估计的数据隐私编程语言