Activity Number:
|
433
- Contributed Poster Presentations: Business Analytics/Statistics Education Interest Group
|
Type:
|
Contributed
|
Date/Time:
|
Wednesday, August 10, 2022 : 10:30 AM to 12:20 PM
|
Sponsor:
|
Business Analytics/Statistics Education Interest Group
|
Abstract #322378
|
|
Title:
|
Predictive and Representative Composite Score
|
Author(s):
|
Heungsun Park* and Seonwoong Eo and Young Rock Kim
|
Companies:
|
Hankuk University of Foreign Studies and Innovation Team, ILHWA Co. LTD and Hankuk University of Foreign Studeis
|
Keywords:
|
composite indicator;
canonical correlation;
best predictor;
nonlinear programing;
canonical correlation coefficient;
principal component analysis
|
Abstract:
|
Composite Score is made up of two or more variables that are highly related to one another conceptually and statistically, and it can be widely used as a predictor for future response variable once developed. This paper aims to develop new composite scoring systems with representing subordinate variables equally as possible whereas principal component analysis/factor analysis fails to do, and being highly correlated with a future variable. The presented methods can be applied to college entrance composite scores taking into account the future better college academic achievement (GPA), or developing the ranking scores for start-up companies using science and technology index scores with keeping in mind the future better stock performances.
|
Authors who are presenting talks have a * after their name.