<?xml version='1.0' encoding='UTF-8'?><?xml-stylesheet href='static/style.xsl' type='text/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>2026-11-22T11:13:51Z</responseDate><request verb="GetRecord" identifier="oai:ecommons.cornell.edu:1813/103331" metadataPrefix="dim">https://ecommons.cornell.edu/server/oai/request</request><GetRecord><record><header><identifier>oai:ecommons.cornell.edu:1813/103331</identifier><datestamp>2026-05-15T19:42:46Z</datestamp><setSpec>com_1813_35</setSpec><setSpec>col_1813_47</setSpec></header><metadata><dim:dim xmlns:dim="http://www.dspace.org/xmlns/dspace/dim" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:doc="http://www.lyncode.com/xoai" xsi:schemaLocation="http://www.dspace.org/xmlns/dspace/dim http://www.dspace.org/schema/dim.xsd">
   <dim:field mdschema="dc" element="contributor" qualifier="author">Sheng, Yuzhe</dim:field>
   <dim:field mdschema="dc" element="contributor" qualifier="chair">Zhang, K. Max</dim:field>
   <dim:field mdschema="dc" element="contributor" qualifier="committeeMember">MacMartin, Douglas</dim:field>
   <dim:field mdschema="dc" element="date" qualifier="accessioned">2021-03-15T13:34:53Z</dim:field>
   <dim:field mdschema="dc" element="date" qualifier="available">2021-07-11T06:00:25Z</dim:field>
   <dim:field mdschema="dc" element="date" qualifier="issued">2020-12</dim:field>
   <dim:field mdschema="dc" element="identifier" qualifier="other">ProQuest Submission ID: 11115</dim:field>
   <dim:field mdschema="dc" element="identifier" qualifier="other">ProQuest Publication ID: 28259897</dim:field>
   <dim:field mdschema="dc" element="identifier" qualifier="uri">https://hdl.handle.net/1813/103331</dim:field>
   <dim:field mdschema="dc" element="identifier" qualifier="doi">https://doi.org/10.7298/ysjf-cf83</dim:field>
   <dim:field mdschema="dc" element="identifier" qualifier="bibid">13312118</dim:field>
   <dim:field mdschema="dc" element="description">45 pages</dim:field>
   <dim:field mdschema="dc" element="description" qualifier="abstract">Non-Intrusive Load Monitoring (NILM), also referred to as energy disaggre- gation, is a promising technique to extract appliance-level information (such as heating, cooling, refrigeration, and lighting) from aggregated building-level metering data. NILM has the potential to be a cost-effective way to improve en- ergy efficiency and reduce energy waste as sub-metering individual appliances is expensive and inconvenient. A number of energy disaggregation techniques have been studied in the past, but to achieve a high prediction accuracy, NILM requires measuring both real and reactive powers. Metering both real and re- active power can be expensive and often needs additional data storage. In this thesis, we addressed this gap by introducing an algorithm called Gaussian Pro- cess Change Surface (GPCS) to energy disaggregation in that GPCS only re- quires real power compared to other NILM algorithms. A novel approach was to model the aggregate power and the background appliances as two indepen- dent Gaussian Processes, and the target appliance as the difference between the aggregate and the background. In addition, we developed Gaussian Process, Deep Kernel Learning, and other existing state-of-the-art techniques as compar- isons to the GPCS. We tested the effectiveness of this new method on two in- dependent data sets. Our evaluation demonstrated that GPCS can achieve high prediction accuracy while eliminating the need to use reactive power compared to the existing NILM algorithms.</dim:field>
   <dim:field mdschema="dc" element="language" qualifier="iso">en</dim:field>
   <dim:field mdschema="dc" element="subject">Deep kernel learning</dim:field>
   <dim:field mdschema="dc" element="subject">Energy disaggregation</dim:field>
   <dim:field mdschema="dc" element="subject">Gaussian process</dim:field>
   <dim:field mdschema="dc" element="subject">Gaussian process change surface</dim:field>
   <dim:field mdschema="dc" element="subject">Machine learning</dim:field>
   <dim:field mdschema="dc" element="subject">Non-intrusive load monitoring</dim:field>
   <dim:field mdschema="dc" element="title">NON-INTRUSIVE ENERGY DISAGGREGATION IN NONPARAMETRIC BAYESIAN FRAMEWORK</dim:field>
   <dim:field mdschema="dc" element="type">dissertation or thesis</dim:field>
   <dim:field mdschema="dc" element="relation" qualifier="localuri">https://newcatalog.library.cornell.edu/catalog/13312118</dim:field>
   <dim:field mdschema="dc" element="format" qualifier="mimetype">application/pdf</dim:field>
   <dim:field mdschema="thesis" element="degree" qualifier="discipline">Mechanical Engineering</dim:field>
   <dim:field mdschema="thesis" element="degree" qualifier="grantor">Cornell University</dim:field>
   <dim:field mdschema="thesis" element="degree" qualifier="level">Master of Science</dim:field>
   <dim:field mdschema="thesis" element="degree" qualifier="name">M.S., Mechanical Engineering</dim:field>
   <dim:field mdschema="dcterms" element="license">https://hdl.handle.net/1813/59810</dim:field>
   <dim:field mdschema="dspace" element="entity" qualifier="type">Publication</dim:field>
   <dim:field mdschema="cris" element="virtual" qualifier="collection" authority="https://cornell-ecommons.eks.prod.4science.cloud/handle/1813/47" confidence="600">Cornell Theses and Dissertations</dim:field>
   <dim:field mdschema="cris" element="virtual" qualifier="author">Sheng, Yuzhe</dim:field>
   <dim:field mdschema="cris" element="virtualsource" qualifier="collection">5893a6ea-7af3-41d7-abc6-04bcd26ab5df</dim:field>
   <dim:field mdschema="others" element="access-status">open.access</dim:field>
   <dim:field mdschema="others" element="access-status">open.access</dim:field>
   <dim:field mdschema="cerif" element="openaire" authority="" confidence="-1">&lt;Publication xmlns="https://www.openaire.eu/cerif-profile/1.1/" id="342a2b65-71c1-4474-ab92-34484d4f4361">
	&lt;Type xmlns="https://www.openaire.eu/cerif-profile/vocab/COAR_Publication_Types">http://purl.org/coar/resource_type/c_1843&lt;/Type>
	&lt;Language>en&lt;/Language>
   	&lt;Title>NON-INTRUSIVE ENERGY DISAGGREGATION IN NONPARAMETRIC BAYESIAN FRAMEWORK&lt;/Title>
   	&lt;PublishedIn>
    	&lt;Publication>
      	&lt;/Publication>
   	&lt;/PublishedIn>
   	&lt;PublicationDate>2020-12&lt;/PublicationDate>
   	&lt;DOI>https://doi.org/10.7298/ysjf-cf83&lt;/DOI>
   	&lt;Authors>
      	&lt;Author>
        	&lt;DisplayName>Sheng, Yuzhe&lt;/DisplayName>
         	&lt;Affiliation>
         		&lt;OrgUnit>
         		&lt;/OrgUnit>
         	&lt;/Affiliation>
      	&lt;/Author>
	&lt;/Authors>
   	&lt;Editors>
	&lt;/Editors>
    &lt;Publishers>
        &lt;Publisher>
            &lt;OrgUnit />
        &lt;/Publisher>
    &lt;/Publishers>
    &lt;Keyword>Deep kernel learning&lt;/Keyword>
    &lt;Keyword>Energy disaggregation&lt;/Keyword>
    &lt;Keyword>Gaussian process&lt;/Keyword>
    &lt;Keyword>Gaussian process change surface&lt;/Keyword>
    &lt;Keyword>Machine learning&lt;/Keyword>
    &lt;Keyword>Non-intrusive load monitoring&lt;/Keyword>
   	&lt;Abstract>Non-Intrusive Load Monitoring (NILM), also referred to as energy disaggre- gation, is a promising technique to extract appliance-level information (such as heating, cooling, refrigeration, and lighting) from aggregated building-level metering data. NILM has the potential to be a cost-effective way to improve en- ergy efficiency and reduce energy waste as sub-metering individual appliances is expensive and inconvenient. A number of energy disaggregation techniques have been studied in the past, but to achieve a high prediction accuracy, NILM requires measuring both real and reactive powers. Metering both real and re- active power can be expensive and often needs additional data storage. In this thesis, we addressed this gap by introducing an algorithm called Gaussian Pro- cess Change Surface (GPCS) to energy disaggregation in that GPCS only re- quires real power compared to other NILM algorithms. A novel approach was to model the aggregate power and the background appliances as two indepen- dent Gaussian Processes, and the target appliance as the difference between the aggregate and the background. In addition, we developed Gaussian Process, Deep Kernel Learning, and other existing state-of-the-art techniques as compar- isons to the GPCS. We tested the effectiveness of this new method on two in- dependent data sets. Our evaluation demonstrated that GPCS can achieve high prediction accuracy while eliminating the need to use reactive power compared to the existing NILM algorithms.&lt;/Abstract>
	&lt;Access xmlns="http://purl.org/coar/access_right" 
    >
    &lt;/Access>
&lt;/Publication>
</dim:field>
</dim:dim>
</metadata></record></GetRecord></OAI-PMH>