Activity Number:
|
589
- Topics in Data Mining, Forecasting, and Bayesian Inference for National Security
|
Type:
|
Contributed
|
Date/Time:
|
Wednesday, August 1, 2018 : 2:00 PM to 3:50 PM
|
Sponsor:
|
Section on Statistics in Defense and National Security
|
Abstract #330035
|
Presentation
|
Title:
|
A Solution to the Forensic Identification of Source Problems Using Fiducial Inference
|
Author(s):
|
Danica Ommen* and Jan Hannig and Jonathan Williams
|
Companies:
|
Iowa State University and University of North Carolina and University of North Carolina
|
Keywords:
|
forensics;
fiducial;
Bayesian
|
Abstract:
|
The application of statistics to the field of forensic science is currently under a major reform after many calls to strengthen its foundations. Many voices from Europe are battling for a Bayesian framework of evidence interpretation while receiving push-back from many US voices who prefer a more transparent, classical approach. This power-struggle of the two main statistical paradigms is leaving the door open for other paradigms of statistics to gather support. In this presentation, we will discuss a potential solution to the forensic identification of source problems using a generalized fiducial inference approach. Advances towards developing a general Fiducial Factor for the two commonly encountered forensic identification of source problems, (unknown) common source and (fixed) specific source, will be presented. We will explore relationships to the supported Bayesian approach, and assess the strengths and weaknesses of the Fiducial approach.
|
Authors who are presenting talks have a * after their name.