Page 1 |
Save page Remove page | Previous | 1 of 128 | Next |
|
small (250x250 max)
medium (500x500 max)
large ( > 500x500)
Full Resolution
All (PDF)
|
This page
All
Subset |
GENERALIZED OPTIMAL LOCATION PLANNING
by
Parisa Ghaemi
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(COMPUTER SCIENCE)
December 2012
Copyright 2012 Parisa Ghaemi
Object Description
| Title | Generalized optimal location planning |
| Author | Ghaemi, Parisa |
| Author email | ghaemi@usc.edu;parissa_ghaemi@yahoo.com |
| Degree | Doctor of Philosophy |
| Document type | Dissertation |
| Degree program | Computer Science |
| School | Viterbi School of Engineering |
| Date defended/completed | 2012-10-24 |
| Date submitted | 2012-11-27 |
| Date approved | 2012-11-27 |
| Restricted until | 2012-11-27 |
| Date published | 2012-11-27 |
| Advisor (committee chair) | Wilson, John |
| Advisor (committee member) |
Shahabi, Cyrus Moore, James |
| Abstract | Optimal location queries have been widely used in spatial decision support systems and marketing in recent years. Given a set S of sites and a set O of weighted objects, a ""basic optimal location query"" finds the location(s) where introducing a new site maximizes the total weight of the objects that are closer to the new site than to any other site. Due to the intrinsic computational complexity of the optimal location problem, researchers have often resorted to making simplifying assumptions in order for the proposed solutions to scale with large datasets. However, there are many real-world applications where such restrictive assumptions may not hold. In this dissertation, we relax three of the aforementioned simplifying assumptions and correspondingly propose solutions for three popular variations of the basic optimal location problem, namely the ""optimal network location problem"", the ""multi-criteria optimal location problem"" and the ""dynamic optimal location problem"". These variations of the original problem allow for considering network distance (rather than p-norm distance), multiple preference criteria (rather than distance as the single preference criterion), and dynamic objects and sites (rather than static ones), respectively. In Chapter 3, we introduce two complementary approaches for efficient computation of optimal network location (ONL) queries, namely EONL (short for ""Expansion-based ONL"") and BONL (short for ""Bound-based ONL""), which enable efficient computation of ONL queries with object-datasets containing uniform and skewed distributions, respectively. Thereafter, we experimentally compare our proposed approaches and discuss their use cases with different real-world applications. Our experimental results with real datasets show that given uniformly distributed object-datasets (i.e., datasets with uniform spatial distributions), EONL is an order of magnitude faster than BONL, whereas with object-datasets with skewed distributions BONL outperforms EONL. Therefore, EONL and BONL have their own exclusive use cases in real-world applications and are complementary. ❧ In Chapter 4, we formalize the multi-criteria location problem as maximal reverse skyline query (MaxRSKY) and introduce two filter-and-refine approaches termed ""Basic-Filtering"" and ""Grid-based-Filtering"" that allow for efficient computation of MaxRSKY queries. The latter approach is an enhanced solution because it avoids redundant computation by filtering out the irrelevant parts of the search space for improved efficiency. Our extensive empirical analysis with both real-world and synthetic datasets show that our enhanced solution is more efficient in computing answers for MaxRSKY queries with large datasets containing thousands of sites and objects. For the datasets that the ""Basic-Filtering"" approach responds to a MaxRSKY query in hours, this computation can be completed in minutes using the ""Grid-based-Filtering"" approach. In Chapter 5, we formalize dynamic optimal network location queries as Continuous Maximal Reverse Nearest Neighbor (CMaxRNN) queries on Spatial Networks, and present a scalable and exact solution for CMaxRNN query computation. In our proposed approach we avoid computation of the optimal location query from scratch, and instead, compute the query incrementally to leverage computations from past queries. Our experimental results on a real-world dataset shows that the CMAxRNN queries are about two orders of magnitude faster than running the optimal location query from scratch. |
| Keyword | optimal location queries; optimal network location queries; multi-criteria optimal location queries; dynamic optimal location queries; spatial networks; maximum reverse nearest neighbor queries; maximum reverse skyline queries; continuous maximal reverse nearest neighbor queries |
| Language | English |
| Part of collection | University of Southern California dissertations and theses |
| Publisher (of the original version) | University of Southern California |
| Place of publication (of the original version) | Los Angeles, California |
| Publisher (of the digital version) | University of Southern California. Libraries |
| Provenance | Electronically uploaded by the author |
| Type | texts |
| Legacy record ID | usctheses-m |
| Rights | Ghaemi, Parisa |
| Access conditions | The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given. |
| Repository name | University of Southern California Digital Library |
| Repository address | USC Digital Library, University of Southern California, University Park Campus MC 7002, 106 University Village, Los Angeles, California 90089-7002, USA |
| Repository email | cisadmin@usc.edu |
| Archival file | uscthesesreloadpub_Volume6/etd-GhaemiPari-1353.pdf |
Description
| Title | Page 1 |
| Full text | GENERALIZED OPTIMAL LOCATION PLANNING by Parisa Ghaemi A Dissertation Presented to the FACULTY OF THE USC GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (COMPUTER SCIENCE) December 2012 Copyright 2012 Parisa Ghaemi |
Comments
Post a Comment for Page 1
