The views expressed here are those of the individual authors and not necessarily those of the JSM sponsors, their officers, or their staff.
Online Program Home
Abstract Details
|
Activity Number:
|
506
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 1, 2012 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Risk Analysis
|
| Abstract - #306146 |
|
Title:
|
Grouping Variables and Collapsing Levels in Credit Scoring
|
|
Author(s):
|
Yoshinori Kawasaki*+ and Masao Ueki
|
|
Companies:
|
Institute of Statistical Mathematics and Yamagata University
|
|
Address:
|
10-3 Midori-Cho, Tokyo, International, 1908562, Japan
|
|
Keywords:
|
Automatic grouping ;
Lasso ;
Smooth-thresholding ;
Variable selection ;
Credit scoring
|
|
Abstract:
|
In building a credit scoring model, we often find some variables having comparable explanatory power. In such a case, it is useful to consider overlapping parameters of which the values are similar. The problem is how to group the variables systematically. We may naturally consider interaction terms. If it happens that all the interaction terms from two categorical variables X and Y have similar effects, then we collapse the conditioning on Y and use X only. This is again the problem of grouping the variables. This paper illustrates how the smooth-thresholding estimating equation (STEE) method works well in automatic variable grouping. STEE simultaneously enables variable selection that can yield sparse solution. It also enjoys the oracle property while does not involve any convex optimization.
|
The address information is for the authors that have a + after their name.
Authors who are presenting talks have a * after their name.
Back to the full JSM 2012 program
|
2012 JSM Online Program Home
For information, contact jsm@amstat.org or phone (888) 231-3473.
If you have questions about the Continuing Education program, please contact the Education Department.