权益分类	功能权益	普通用户	{{item.name}}会员
{{category.name}}	{{benefitItem.name}}

BIGDATA: Mid-Scale: DA: Analytical Approaches to Massive Data Computation with Applications to Genomics

BIGDATA：中型：DA：海量数据计算的分析方法及其在基因组学中的应用

基本信息

批准号：
1247581
负责人：
Eli Upfal
金额：
$ 156.67万
依托单位：
Brown University
依托单位国家：
美国
项目类别：
Standard Grant
财政年份：
2012
资助国家：
美国
起止时间：
2012-10-01 至 2018-09-30
项目状态：
已结题

来源：
https://www.nsf.gov/awardsearch/showAward?AWD_ID=1247581&HistoricalAwards=false
关键词：
BIGDATA Mid Scale DA Analytical

项目摘要

The goal of this project is to design and test mathematically well-founded algorithmic and statistical techniques for analyzing large scale, heterogeneous and noisy data. The proposed research is transformative in its emphasis on rigorous analytical evaluation of algorithms' performance and statistical measures of output uncertainty, in contrast to the primarily heuristic approaches currently used in data mining and machine learning. Any progress in that direction will have significant contribution to the reliability and scientific impact of massive data analysis. This project is motivated by the challenges in analyzing molecular biology data. Molecular biology provides an excellent source of data for testing advanced data analysis techniques: specifically, DNA/RNA sequence data repositories are growing at a super-exponential rate. The data is typically large and noisy, and in some cases includes both genotype and phenotype features that permits experimental validation of the analysis. However, the methods and techniques developed in this project will be broadly applicable to other scientific communities that process massive multi-variant data sets. The major technical goals of the project include: (1) Design efficient algorithms that provide guarantees on the output when the data comes from independent random samples from an unknown distribution. (2) Develop techniques for estimating the minimum number of samples required to test hypothesis of varying complexity in large datasets, building on techniques in computational statistics. (3) Design algorithms to analyze data on graphs that represent interactions between samples or features in the dataset. These data may be static (e.g. mutations on interacting genes represented by a protein interaction network) or dynamic (e.g. information dissemination on a social network).This project will advocate a responsible approach to data analysis, based on well-founded mathematical and statistical concepts. The capacity building activities of the project include: (1) Creation and dissemination of algorithms and software that implement rigorous computational and statistical approaches to big data analysis. (2) Educational initiatives at the graduate and undergraduate level to build a bigger workforce of data scientists with the appropriate foundational skills both to apply analytical tools to existing datasets and to develop new approaches to future datasets. The proposed work will be tested on extensive cancer genome data, contributing to health IT, one of the National Priority Domain Areas.

该项目的目标是设计和测试数学上有充分依据的算法和统计技术，用于分析大规模，异构和噪声数据。与目前在数据挖掘和机器学习中使用的主要启发式方法相比，所提出的研究在其对算法性能的严格分析评估和输出不确定性的统计测量方面具有变革性。这方面的任何进展都将对大规模数据分析的可靠性和科学影响作出重大贡献。该项目的动机是分析分子生物学数据的挑战。分子生物学为测试先进的数据分析技术提供了极好的数据来源：具体而言，DNA/RNA序列数据库正以超指数的速度增长。数据通常是大的和嘈杂的，并且在某些情况下包括允许分析的实验验证的基因型和表型特征。然而，在这个项目中开发的方法和技术将广泛适用于处理大量多变量数据集的其他科学界。该项目的主要技术目标包括：（1）设计有效的算法，当数据来自未知分布的独立随机样本时，该算法提供对输出的保证。(2)在计算统计学技术的基础上，开发用于估计在大型数据集中检验不同复杂性假设所需的最低样本数的技术。(3)设计算法来分析表示数据集中样本或特征之间交互的图形上的数据。这些数据可能是静态的（例如蛋白质相互作用网络所代表的相互作用基因的突变）或动态的（例如社交网络上的信息传播）。该项目将倡导基于有根据的数学和统计概念的负责任的数据分析方法。该项目的能力建设活动包括：（1）创建和传播算法和软件，以实施严格的计算和统计方法进行大数据分析。 (2)在研究生和本科阶段的教育计划，以建立一个更大的数据科学家的劳动力与适当的基础技能，既应用分析工具，以现有的数据集和开发新的方法，以未来的数据集。拟议的工作将在广泛的癌症基因组数据上进行测试，为国家优先领域之一的健康信息技术做出贡献。