Grant: Discovery Grants Program - Individual
University of Waterloo
$60,000.00
Adaptive Data Mining Algorithms for Non-Stationary Data
FY 2015-16
Summary
Today's information and communication systems are characterized by the continuous generation, transfer and storage of massive amounts of data. Most of these data typically evolve over time and their properties change from one domain to another. Although different data mining techniques have been proposed for analyzing large-scale data, this non-stationary nature of the data negatively affects the sustainability of these techniques over time and prevents their applicability to new domains. For instance, systems for product recommendations cannot quickly adapt to the ever-changing needs of the customers. These shortcomings motivate the research and development of data mining algorithms that can quickly adapt to changes in non-stationary data.
The objective of this research is to provide scalable adaptive algorithms for non-stationary data mining. The main characteristics of these algorithms are adaptability and scalability. Adaptability is to cope with the nature of non-stationary data and provide a type of personalized data mining. Scalability is to deal with the sustainability of these algorithms in handling large volumes of data which are a dominant characteristic in many of today's applications. The proposed research spans different aspects of data mining techniques such as classification, clustering, and novelty detection. The proposed research is structured into three major research tasks. The first task involves investigating the development of supervised learning techniques which adapt to new domains while alleviating the complexities of re-building learning models from scratch. The second task explores the adaption of unsupervised learning algorithms such as clustering and novelty detection. Specifically, design and development of adaptive algorithms for similarity-based clustering which quickly and effectively update the existing clusters when new data are encountered will be pursued. Novel detection algorithms for the detection of evolving and emerging concepts in the new data will be developed. In the third task we will investigate adaptability in personalized learning where the existence of a different of users generates a challenging aspect of non-stationarity in the data. In this problem, the very few data associated with a new user or group of users will be used to adapt the learning models at hand to perform well on these data.
The proposed research will advance the state-of-the-art in data mining and machine learning algorithms for large scale non-stationary data. The outcome of this research has practical importance and will provide data mining solutions and highly qualified personnel with skills that will benefit several application sectors that demand advanced data mining technology to cope with dynamic and evolving data. The results of this research have strong potential for commercialization into products for personalized and adaptive data mining
Program
Discovery Grants Program - Individual
Recipient
Location
Ontario, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data
As published
Application id
RGPIN-2015-05224
Name nom
Kamel, Mohamed
Department département
Electrical & Computer Engineering - Electrical & Computer Engineering
Organization id
33
Institution établissement
University of Waterloo
Province en
Ontario
Province fr
Ontario
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2015
Competition year année de concours
2015
Award amount
60000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
Discovery grants programs
Group fr
Programmes de subventions à la découverte
Committee code
1507
Committee name en
Computer Science
Committee name fr
Sciences informatiques
Area of application code
1200
Area of application group en
Advancement of knowledge
Area of application group fr
Avancement des connaissances
Area of application en
Advancement of knowledge
Area of application fr
Avancement des connaissances
Research subject code
2805
Research subject group en
Artificial intelligence
Research subject group fr
Intelligence artificielle
Research subject en
Learning and inference theories
Research subject fr
Théories de l'apprentissage et l'inférence
Partie
1
Nb partie
5
Application title
Adaptive Data Mining Algorithms for Non-Stationary Data
Keyword
NSERC
Application summary
Today's information and communication systems are characterized by the continuous generation, transfer and storage of massive amounts of data. Most of these data typically evolve over time and their properties change from one domain to another. Although different data mining techniques have been proposed for analyzing large-scale data, this non-stationary nature of the data negatively affects the sustainability of these techniques over time and prevents their applicability to new domains. For instance, systems for product recommendations cannot quickly adapt to the ever-changing needs of the customers. These shortcomings motivate the research and development of data mining algorithms that can quickly adapt to changes in non-stationary data.<br />
The objective of this research is to provide scalable adaptive algorithms for non-stationary data mining. The main characteristics of these algorithms are adaptability and scalability. Adaptability is to cope with the nature of non-stationary data and provide a type of personalized data mining. Scalability is to deal with the sustainability of these algorithms in handling large volumes of data which are a dominant characteristic in many of today's applications. The proposed research spans different aspects of data mining techniques such as classification, clustering, and novelty detection. The proposed research is structured into three major research tasks. The first task involves investigating the development of supervised learning techniques which adapt to new domains while alleviating the complexities of re-building learning models from scratch. The second task explores the adaption of unsupervised learning algorithms such as clustering and novelty detection. Specifically, design and development of adaptive algorithms for similarity-based clustering which quickly and effectively update the existing clusters when new data are encountered will be pursued. Novel detection algorithms for the detection of evolving and emerging concepts in the new data will be developed. In the third task we will investigate adaptability in personalized learning where the existence of a different of users generates a challenging aspect of non-stationarity in the data. In this problem, the very few data associated with a new user or group of users will be used to adapt the learning models at hand to perform well on these data.<br />
The proposed research will advance the state-of-the-art in data mining and machine learning algorithms for large scale non-stationary data. The outcome of this research has practical importance and will provide data mining solutions and highly qualified personnel with skills that will benefit several application sectors that demand advanced data mining technology to cope with dynamic and evolving data. The results of this research have strong potential for commercialization into products for personalized and adaptive data mining<br />
<br />
<p>
</p>