<?xml version="1.0" encoding="UTF-8"?><?xml-stylesheet type="text/xsl" href="static/style.xsl"?><OAI-PMH xmlns="http://www.openarchives.org/OAI/2.0/" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.openarchives.org/OAI/2.0/ http://www.openarchives.org/OAI/2.0/OAI-PMH.xsd"><responseDate>2026-11-22T01:42:46Z</responseDate><request verb="GetRecord" identifier="oai:uwspace.uwaterloo.ca:10012/3000" metadataPrefix="dim">https://dspacemainprd01.lib.uwaterloo.ca/server/oai/request</request><GetRecord><record><header><identifier>oai:uwspace.uwaterloo.ca:10012/3000</identifier><datestamp>2017-03-21T19:36:23Z</datestamp><setSpec>com_10012_9900</setSpec><setSpec>com_10012_1</setSpec><setSpec>col_10012_9908</setSpec><setSpec>col_10012_6</setSpec></header><metadata><dim:dim xmlns:dim="http://www.dspace.org/xmlns/dspace/dim" xmlns:doc="http://www.lyncode.com/xoai" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.dspace.org/xmlns/dspace/dim http://www.dspace.org/schema/dim.xsd">
   <dim:field mdschema="dc" element="contributor" qualifier="author" authority="d21d0d01-68d6-40eb-85c2-b8a5d849edd6" confidence="600">Sun, Yanmin</dim:field>
   <dim:field mdschema="dc" element="date" qualifier="accessioned">2007-05-11T17:57:27Z</dim:field>
   <dim:field mdschema="dc" element="date" qualifier="available">2007-05-11T17:57:27Z</dim:field>
   <dim:field mdschema="dc" element="date" qualifier="issued">2007-05-11T17:57:27Z</dim:field>
   <dim:field mdschema="dc" element="date" qualifier="submitted">2007-05-11</dim:field>
   <dim:field mdschema="dc" element="identifier" qualifier="uri">http://hdl.handle.net/10012/3000</dim:field>
   <dim:field mdschema="dc" element="description" qualifier="abstract" lang="en">The classification of data with imbalanced class distributions has&#xd;
posed a significant drawback in the performance attainable by most&#xd;
well-developed classification systems, which assume relatively&#xd;
balanced class distributions. This problem is especially crucial&#xd;
in many application domains, such as medical diagnosis, fraud&#xd;
detection, network intrusion, etc., which are of great importance&#xd;
in machine learning and data mining.&#xd;
&#xd;
This thesis explores meta-techniques which are applicable to most&#xd;
classifier learning algorithms, with the aim to advance the&#xd;
classification of imbalanced data. Boosting is a powerful&#xd;
meta-technique to learn an ensemble of weak models with a promise&#xd;
of improving the classification accuracy. AdaBoost has been taken&#xd;
as the most successful boosting algorithm. This thesis starts with&#xd;
applying AdaBoost to an associative classifier for both learning&#xd;
time reduction and accuracy improvement. However, the promise of&#xd;
accuracy improvement is trivial in the context of the class&#xd;
imbalance problem, where accuracy is less meaningful. The insight&#xd;
gained from a comprehensive analysis on the boosting strategy of&#xd;
AdaBoost leads to the investigation of cost-sensitive boosting&#xd;
algorithms, which are developed by introducing cost items into the&#xd;
learning framework of AdaBoost. The cost items are used to denote&#xd;
the uneven identification importance among classes, such that the&#xd;
boosting strategies can intentionally bias the learning towards&#xd;
classes associated with higher identification importance and&#xd;
eventually improve the identification performance on them. Given&#xd;
an application domain, cost values with respect to different types&#xd;
of samples are usually unavailable for applying the proposed&#xd;
cost-sensitive boosting algorithms. To set up the effective cost&#xd;
values, empirical methods are used for bi-class applications and&#xd;
heuristic searching of the Genetic Algorithm is employed for&#xd;
multi-class applications.&#xd;
&#xd;
&#xd;
This thesis also covers the implementation of the proposed&#xd;
cost-sensitive boosting algorithms. It ends with a discussion on&#xd;
the experimental results of classification of real-world&#xd;
imbalanced data. Compared with existing algorithms, the new&#xd;
algorithms this thesis presents are superior in achieving better&#xd;
measurements regarding the learning objectives.</dim:field>
   <dim:field mdschema="dc" element="format" qualifier="extent">739318 bytes</dim:field>
   <dim:field mdschema="dc" element="format" qualifier="mimetype">application/pdf</dim:field>
   <dim:field mdschema="dc" element="language" qualifier="iso" lang="en">en</dim:field>
   <dim:field mdschema="dc" element="publisher" lang="en">University of Waterloo</dim:field>
   <dim:field mdschema="dc" element="subject" lang="en">Classification</dim:field>
   <dim:field mdschema="dc" element="subject" lang="en">Imbalanced Data</dim:field>
   <dim:field mdschema="dc" element="subject" lang="en">Cost-Sensitive Learning</dim:field>
   <dim:field mdschema="dc" element="subject" lang="en">Boosting</dim:field>
   <dim:field mdschema="dc" element="subject" qualifier="program" lang="en">Electrical and Computer Engineering (Software Engineering)</dim:field>
   <dim:field mdschema="dc" element="title" lang="en">Cost-Sensitive Boosting for Classification of Imbalanced Data</dim:field>
   <dim:field mdschema="dc" element="type" lang="en">Doctoral Thesis</dim:field>
   <dim:field mdschema="dc" element="pending" lang="en">false</dim:field>
   <dim:field mdschema="uws-etd" element="degree" qualifier="department" lang="en">Electrical and Computer Engineering</dim:field>
   <dim:field mdschema="uws-etd" element="degree" lang="en">Doctor of Philosophy</dim:field>
   <dim:field mdschema="uws" element="typeOfResource" lang="en">Text</dim:field>
   <dim:field mdschema="uws" element="peerReviewStatus" lang="en">Unreviewed</dim:field>
   <dim:field mdschema="uws" element="scholarLevel" lang="en">Graduate</dim:field>
   <dim:field mdschema="others" element="access-status">open.access</dim:field>
</dim:dim>
</metadata></record></GetRecord></OAI-PMH>