Grant: Discovery Grants Program - Individual
University of Manitoba
$12,000.00
Classification methodology for mitigating the effects of unreliable class assignments in high dimensional biomedical data
FY 2008-09
Summary
Classifying biomedical data involves finding a mapping (relationship) from patterns (e.g., data relating to some type of tissue) to a set of classes (e.g., disease states). Patterns are represented by features (e.g., concentrations of biological compounds) and class labels are assigned using a reference test (e.g., a medical expert's analysis of tissue being "normal" or "abnormal"). This process often suffers from three significant challenges: the number of features in a pattern is high; the number of patterns is low; and the reference test may be unreliable. The first two challenges, known collectively as the "curse of dimensionality", cause an inability to find robust, general solutions. This is often addressed by reducing, in some fashion, the number of features; however, a direct correspondence back to the original features is necessary for medical experts to make informed judgments about the mapping's predictive power. While reference tests may be well-established benchmarks, they are seldom perfectly accurate and sometimes improperly applied. Nevertheless, any strategy that compensates for this imprecision must ensure that the mapping is correctly validated against the benchmark.
Techniques using Computational Intelligence (CI) - adaptive, learning, or evolutionary algorithms - have demonstrated their effectiveness in data analysis. This proposal involves the development of a comprehensive CI based Classification Methodology to mitigate the effects relating to many features, few patterns, and an unreliable reference test. Its novelty and significance is in its tightly integrated approach, where the enterprises - accurate classification, effective feature reduction, unreliable class assignment mitigation - incrementally inform and drive each other in a unified manner. This methodology will be instrumental in the analysis of complex biomedical data and will lead to more reliable classification systems. Further, while this investigation involves methodological and foundational concepts, it will simultaneously deal with "real-world" data. This synchronization of the fundamentals with their use in the specific application domain will ensure an efficacious and coherent comprehension of the underlying processes and will generate tangible practical outcomes.
Program
Discovery Grants Program - Individual
Recipient
Location
Manitoba, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data
As published
Application id
217281-2008
Name nom
Pizzi, Nicolino
Department département
Computer Science
Organization id
19
Institution établissement
University of Manitoba
Province en
Manitoba
Province fr
Manitoba
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2008
Competition year année de concours
2008
Award amount
12000
Program id
RGPIN
Program nane 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
331
Committee name en
Computing and Information Sciences - B
Committee name fr
Informatique - B
Area of application code
1207
Area of application group en
Advancement of knowledge
Area of application group fr
Avancement des connaissances
Area of application en
Information, computer and communication technologies
Area of application fr
Technologie de l'information, des ordinateurs et des réseaux de communications
Research subject code
2605
Research subject group en
Robotics
Research subject group fr
Robotique
Research subject en
Pattern analysis and machine intelligence
Research subject fr
Analyse de formes et intelligence artificielle
Partie
1
Nb partie
1
Application title
Classification methodology for mitigating the effects of unreliable class assignments in high dimensional biomedical data
Keyword
NA
Application summary
Classifying biomedical data involves finding a mapping (relationship) from patterns (e.g., data relating to some type of tissue) to a set of classes (e.g., disease states). Patterns are represented by features (e.g., concentrations of biological compounds) and class labels are assigned using a reference test (e.g., a medical expert's analysis of tissue being "normal" or "abnormal"). This process often suffers from three significant challenges: the number of features in a pattern is high; the number of patterns is low; and the reference test may be unreliable. The first two challenges, known collectively as the "curse of dimensionality", cause an inability to find robust, general solutions. This is often addressed by reducing, in some fashion, the number of features; however, a direct correspondence back to the original features is necessary for medical experts to make informed judgments about the mapping's predictive power. While reference tests may be well-established benchmarks, they are seldom perfectly accurate and sometimes improperly applied. Nevertheless, any strategy that compensates for this imprecision must ensure that the mapping is correctly validated against the benchmark.
Techniques using Computational Intelligence (CI) - adaptive, learning, or evolutionary algorithms - have demonstrated their effectiveness in data analysis. This proposal involves the development of a comprehensive CI based Classification Methodology to mitigate the effects relating to many features, few patterns, and an unreliable reference test. Its novelty and significance is in its tightly integrated approach, where the enterprises - accurate classification, effective feature reduction, unreliable class assignment mitigation - incrementally inform and drive each other in a unified manner. This methodology will be instrumental in the analysis of complex biomedical data and will lead to more reliable classification systems. Further, while this investigation involves methodological and foundational concepts, it will simultaneously deal with "real-world" data. This synchronization of the fundamentals with their use in the specific application domain will ensure an efficacious and coherent comprehension of the underlying processes and will generate tangible practical outcomes.