|
Activity Number:
|
478
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 1, 2007 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistical Computing
|
| Abstract - #309985 |
|
Title:
|
Rule and Rule Ensemble-Based Methods for Financial Crime Detection
|
|
Author(s):
|
Fernando Cela Diaz*+ and Agus Sudjianto
|
|
Companies:
|
Bank of America and Bank of America
|
|
Address:
|
200 College St, Charlotte, NC, 28255,
|
|
Keywords:
|
Financial Crime ; Fraud ; Machine Learning ; Data Mining ; Rule ; Rule Ensembles
|
|
Abstract:
|
Rule inference methods are commonly used in Financial Crime Detection. Both supervised and unsupervised methods are possible. Recently, the idea of rule ensembles has been proposed as a method for improving the performance of weak rule learners. This paper will present existing methodologies and explore current real-world applications with emphasis on empirical performance evaluation in real settings.
|
