| EPSRC Reference: |
GR/S73631/01 |
| Title: |
The use of combinatorial clustering to develop predictive risk models of disease from observational epidemiological data |
| Principal Investigator: |
Morgan, Professor KL |
| Other Investigators: |
|
| Researcher Co-Investigators: |
|
| Project Partners: |
|
| Department: |
Veterinary Clinical Science |
| Organisation: |
University of Liverpool |
| Scheme: |
Standard Research (Pre-FEC) |
| Starts: |
14 April 2004 |
Ends: |
13 September 2004 |
Value (£): |
6,177
|
| EPSRC Research Topic Classifications: |
| Logic & Combinatorics |
Population Ecology |
| Statistics & Appl. Probability |
|
|
| EPSRC Industrial Sector Classifications: |
| Healthcare |
Pharmaceuticals and Biotechnology |
|
| Related Grants: |
|
| Panel History: |
|
|
Summary on Grant Application Form |
|
THIS IS A PROPOSAL FOR AN OVERSEAS TRAVEL GRANT. ITS OBJECTIVE IS TO FACILITATE AN INTERDISCIPLINARY PILOT RESEARCH PROJECT. THE STRATEGIC AIM OF THIS IS TO INITIATE AND ENABLE THE DEVELOPMENT OF A COLLABORATIVE RESEARCH PROGRAMME INVOLVING COMPUTER SCIENTISTS AND EPIDEMIOLOGISTS.ITS SPECIFIC FOCUS IS A PROBLEM ENCOUNTERED IN THE ANALYSIS OF ALL OBSERVATIONAL EPIDEMIOLOGICAL DATA. HOW BEST TO IDENTIFY THE VARIABLES ASSOCIATED WITH A DISEASE (RISK FACTORS) FROM THE LARGE NUMBER OF VARIABLES FOR WHICH DATA ARE COLLECTED AND HOW TO DEAL WITH POTENTIAL INTERACTIONS BETWEEN THEM.CURRENT METHODS INVOLVE THE USE OF MULTIVARIABLE TECHNIQUES, COMMONLY LOGISTIC REGRESSION. THIS ALLOWS THE EFFECT OF A VARIABLE TO BE MEASURED WHEN ADJUSTED FOR THE EFFECTS OF ALL OTHER VARIABLES. A MAJOR PROBLEM WITH THIS TECHNIQUE IS THAT THE NUMBER OF VARIABLES WHICH CAN BE REALISTICALLY INCLUDED IN THE MODEL IS LIMITED. CONSEQUENTLY, UNIVARIABLE SCREENING METHODS,SOMETIMES MODULATED BY SUBJECTIVE ASSESSMENTS, ARE USED TO REDUCE THE NUMBER OF VARIABLES PRIOR TO THE DEVELOPMENT OF LOGISTIC REGRESSION MODELS.THIS PROPOSAL WILL FOCUS ON ELIMINATING THE RESTRICTIONS RELATED TO THE SIZE OF EPIDEMIOLOGICAL DATA MENTIONED ABOVE AS WELL AS ON THE PRECISION OF THE RESULTS. THE MAIN IDEA IS TO HAVE A CARDINAL CHANGE IN INFORMATION TECHNOLOGY SUPPORTING THE WHOLE PROCESS OF DATA ANALYSIS. WE PROPOSE TO ANALYSE OBSERVATIONAL EPIDEMIOLOGICAL DATA BY LINKING COMBINATIONAL CLUSTERING AND SUPPORT VECTOR MACHINE LEARNING METHODS TO CONSTRUCT PREDICTION MODELS. THE MODELS WILL BE USED NOT ONLY AS PREDICTIVE TOOLS BUT ALSO AS EFFICIENT SYSTEMS TO ESTIMATE THE SIGNIFICANCE OF VARIABLES BY, FOR EXAMPLE, CALCULATING INTEGRAL INDICES AND GROUPING OBSERVATIONS INTO CLUSTERS. THIS PROPOSAL IS THE FIRST STEP IN THE MODERNISATION OF COMPUTATIONAL METHODS FOR EPIDEMIOLOGICAL PURPOSES THE RESULTS OF THIS COLLABORATION WILL BE PUBLISHED IN SCIENTIFIC LITERATURE AND WILL FORM THE BASIS OF FUTURE COLLABORATIVE RESEARCH PROPOSALS
|
|
Key Findings |
|
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
|
|
Potential use in non-academic contexts |
|
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
|
|
Impacts |
|
| Description |
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk |
| Summary |
|
| Date Materialised |
|
|
|
|
Sectors submitted by the Researcher |
|
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
|
| Project URL: |
|
| Further Information: |
|
| Organisation Website: |
http://www.liv.ac.uk |