Model Distribution Effects on Likelihood Ratios in Fire Debris Analysis

• NCJ Number: 253121
• Journal: Separations Volume: 5 Issue: 3 Dated: 2018
• Author(s): Allysa Allen; Mary R. Williams; Nickolas A. Thurn; Michael E. Sigman
• Date Published: 2018
• Length: 24 pages
• Grant Number(s): 2015-DN-BX-K051
• Sponsoring Agency: National Institute of Justice (NIJ)
• Publication Format: Article
• Publication Type: Research (Applied/Empirical); Report (Study/Research); Report (Grant Sponsored); Program/Project Description
• Language: English
• Country: United States of America

Annotation

This project developed computational models for determining the strength of fire debris evidence based on likelihood ratios (LR) were developed and validated against data sets derived from different distributions of ASTM E1618-14 designated ignitable liquid class and substrate pyrolysis contributions, using in-silico generated data.

Abstract

The models all performed well in cross validation against the distributions used to generate the model; however, a model generated based on data that does not contain representatives from all of the ASTM E1618-14 classes did not perform well in validation with data sets that contain representatives from the missing classes. A quadratic discriminant model based on a balanced data set (ignitable liquid versus substrate pyrolysis), with a uniform distribution of the ASTM E1618-14 classes, performed well (receiver operating characteristic area under the curve of 0.836) when tested against laboratory-developed, casework-relevant samples of known ground truth. (publisher abstract modified)