|
Activity Number:
|
271
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 4, 2009 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statistical Graphics
|
| Abstract - #304241 |
|
Title:
|
Tools for Identifying Homogenous Subgroups in Large Data
|
|
Author(s):
|
Yuanyuan Huang*+ and Heike Hofmann and Dianne Cook
|
|
Companies:
|
Iowa State University and Iowa State University and Iowa State University
|
|
Address:
|
2426 Howe Hall #4876, Ames, IA, 50011,
|
|
Keywords:
|
data mining ; graphics ; NHANES ; subgroup ; multivariate
|
|
Abstract:
|
With an increase in complexity of data sets the task of finding homogeneous subgroups becomes more challenging and yet more important. The presence of homogeneous subgroups allows a more targeted treatment of individuals in all aspects of life. At the example of the National Health and Nutrition Examination Survey (NHANES), 1999-2002, we will discuss visual tools of identifying subgroups of the population with similar dietary habits. In homogeneous sub-populations with "good" quality diet, the health status of individuals is related to various explanatory factors. This project studies the NHANES III data using machine learning approaches to data mining and knowledge discovery, to identify the sub-population that benefit most from a change to healthier dietary habits. We compare the results from different multivariate models and discuss possible future applications in medicine.
|
