Please use this identifier to cite or link to this item: http://scholarbank.nus.edu.sg/handle/10635/48690
Title: Gaussian Process-Based Decentralized Data Fusion and Active Sensing Agents: Towards Large-Scale Modeling and Prediction of Spatiotemporal Traffic Phenomena
Authors: CHEN JIE
Keywords: Parallel Gaussian Process, Decentralized Data Fusion, Decentralized Active Sensing, Spatiotemporal modeling and prediction, Mobility on Demand
Issue Date: 16-Aug-2013
Source: CHEN JIE (2013-08-16). Gaussian Process-Based Decentralized Data Fusion and Active Sensing Agents: Towards Large-Scale Modeling and Prediction of Spatiotemporal Traffic Phenomena. ScholarBank@NUS Repository.
Abstract: This thesis is dedicated to large-scale modeling and prediction of spatiotemporal urban traffic phenomena. Towards this goal, our proposed approaches rely on a class of Bayesian non-parametric models: Gaussian processes (GP). A novel relational GP is proposed to accurately model spatiotemporal urban traffic phenomena in real world situation. Three novel parallel GPs are proposed to achieve efficient and scalable urban traffic phenomenon prediction given a large phenomenon data, and tested on two large real world datasets. A decentralized algorithm framework: Gaussian process-based decentralized data fusion and active sensing (D2FAS), which includes two novel decentralized data fusion algorithms and one novel decentralized active sensing algorithm, is proposed to exploit active mobile sensors to perform decentralized perception of the spatiotemporal urban traffic phenomena. The D2FAS algorithms are justified on real world datasets for monitoring traffic conditions and sensing/servicing urban mobility demands with active mobile sensors.
URI: http://scholarbank.nus.edu.sg/handle/10635/48690
Appears in Collections:Ph.D Theses (Open)

Show full item record
Files in This Item:
File Description SizeFormatAccess SettingsVersion 
ChenJ.pdf1.26 MBAdobe PDF

OPEN

NoneView/Download

Page view(s)

180
checked on Dec 11, 2017

Download(s)

214
checked on Dec 11, 2017

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.