國立高雄大學圖資館 |

Language: English

Back

Revenue maximization in online auctions.

University of California, Berkeley.

Revenue maximization in online auctions.

Record Type:	Electronic resources : Monograph/item
Title/Author:	Revenue maximization in online auctions.
Author:	Wu, Felix Tien-Shang.
Description:	69 p.
Notes:	Chair: Christos H. Papadimitriou.
Notes:	Source: Dissertation Abstracts International, Volume: 66-10, Section: B, page: 5515.
Contained By:	Dissertation Abstracts International66-10B.
Subject:	Computer Science.
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3190876
ISBN:	9780542344251

Revenue maximization in online auctions.
Wu, Felix Tien-Shang.

Revenue maximization in online auctions. - 69 p.

Chair: Christos H. Papadimitriou.

Thesis (Ph.D.)--University of California, Berkeley, 2005.

In analyzing the revenue properties of such auctions, we show a connection between this problem and that of online learning with expert advice. We demonstrate how to use learning algorithms to design auctions that are asymptotically optimal. In this context, we focus on learning algorithms that use polynomial weight functions to adapt to new information. We show how such algorithms can have a useful combination of worst-case and average-case properties, suggesting that they might be used in place of more commonly studied algorithms based on exponential weight functions. Finally, we extend our results to the study of posted price mechanisms, in which buyers react to dynamically adjusted posted prices, rather than submitting bids to the seller.

ISBN: 9780542344251Subjects--Topical Terms:

212513
Computer Science.

Revenue maximization in online auctions.
LDR:02922nmm _2200289 _450 001 170891
005 20061228142313.5
008 090528s2005 eng d
020 $a 9780542344251
035 $a 00242921
040 $a UnM $c UnM
100 0 $a Wu, Felix Tien-Shang. $3 244922
245 1 0 $a Revenue maximization in online auctions.
300 $a 69 p.
500 $a Chair: Christos H. Papadimitriou.
500 $a Source: Dissertation Abstracts International, Volume: 66-10, Section: B, page: 5515.
502 $a Thesis (Ph.D.)--University of California, Berkeley, 2005.
520 # $a In analyzing the revenue properties of such auctions, we show a connection between this problem and that of online learning with expert advice. We demonstrate how to use learning algorithms to design auctions that are asymptotically optimal. In this context, we focus on learning algorithms that use polynomial weight functions to adapt to new information. We show how such algorithms can have a useful combination of worst-case and average-case properties, suggesting that they might be used in place of more commonly studied algorithms based on exponential weight functions. Finally, we extend our results to the study of posted price mechanisms, in which buyers react to dynamically adjusted posted prices, rather than submitting bids to the seller.
520 # $a In the modern information economy, markets are proliferating, and in such an environment, automated market mechanisms, such as auctions, are becoming more important than ever. Economists have long studied the problem of maximizing the revenue obtained by the seller from an auction. In this dissertation, we study the same problem, but from the computer scientist's perspective: where the economist generally assumes that inputs are independently drawn from a preexisting population, we take the assumption that inputs are controlled by an adversary. At the same time, we adopt the economist's assumption that inputs are held as private information by individuals, who will only reveal information when it is in their best interest to do so.
520 # $a In this model, we study auctions with a novel combination of two properties. First, we assume that the seller has available an unlimited supply of the good to be sold, as would be the case in auctions for digital goods such as music or movies. Second, we assume that the auction is online, in the sense that it proceeds over time, with the seller forced to make a sequence of decisions, each based only on information received up to that point.
590 $a School code: 0028.
650 # 0 $a Computer Science. $3 212513
650 # 0 $a Economics, Theory. $3 212740
690 $a 0511
690 $a 0984
710 0 # $a University of California, Berkeley. $3 212474
773 0 # $g 66-10B. $t Dissertation Abstracts International
790 $a 0028
790 1 0 $a Papadimitriou, Christos H., $e advisor
791 $a Ph.D.
792 $a 2005
856 4 0 $u http://libsw.nuk.edu.tw:81/login?url=http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3190876 $z http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3190876