Computing with Distributed Information / Yang Li.

Format:

Book

Thesis/Dissertation

Author/Creator:

Li, Yang, author.

Contributor:

Loo, Boon Thau, degree supervisor.

Khanna, Sanjeev, degree supervisor.

Zhang, Qin, degree committee member.

Roth, Aaron, degree committee member.

Kearns, Michael, degree committee member.

Kannan, Sampath, degree committee member.

University of Pennsylvania. Computer and Information Science, degree granting institution.

Language:

English

Subjects (All):

Computer science.

Information science.

Computer and Information Science--Penn dissertations.

Penn dissertations--Computer and Information Science.

Local Subjects:

Computer science.

Information science.

Computer and Information Science--Penn dissertations.

Penn dissertations--Computer and Information Science.

Genre:

Academic theses.

Physical Description:

1 online resource (168 pages)

Contained In:

Dissertation Abstracts International 79-01B(E).

Place of Publication:

[Philadelphia, Pennsylvania]: University of Pennsylvania ; Ann Arbor : ProQuest Dissertations & Theses, 2017.

Language Note:

English

System Details:

Mode of access: World Wide Web.

text file

Summary:

The age of computing with massive data sets is highlighting new computational challenges. Nowadays, a typical server may not be able to store an entire data set, and thus data is often partitioned and stored on multiple servers in a distributed manner. A natural way of computing with such distributed data is to use distributed algorithms: these are algorithms where the participating parties (i.e., the servers holding portions of the data) collaboratively compute a function over the entire data set by sending (preferably small-size) messages to each other, where the computation performed at each participating party only relies on the data possessed by it and the messages received by it.

We study distributed algorithms focused on two key themes: convergence time and data summarization. Convergence time measures how quickly a distributed algorithm settles on a globally stable solution, and data summarization is the approach of creating a compact summary of the input data while retaining key information. The latter often leads to more efficient computation and communication. The main focus of this dissertation is on design and analysis of distributed algorithms for important problems in diverse application domains centering on the themes of convergence time and data summarization. Some of the problems we study include convergence time of double oral auction and interdomain routing, summarizing graphs for large-scale matching problems, and summarizing data for query processing.

Notes:

Source: Dissertation Abstracts International, Volume: 79-01(E), Section: B.

Advisors: Sanjeev Khanna; Boon Thau Loo; Committee members: Sampath Kannan; Michael Kearns; Aaron Roth; Qin Zhang.

Department: Computer and Information Science.

Ph.D. University of Pennsylvania 2017.

Local Notes:

School code: 0175

ISBN:

9780355129656

Access Restriction:

Restricted for use by site license.