Objectifying evidence evaluation for gunshot residue comparisons using machine learning on criminal case data. (June 2022)

• Record Type: Journal Article
• Authors: Matzen, Timo; Kukurin, Corina; van de Wetering, Judith; Ariëns, Simone; Bosma, Wauter; Knijnenberg, Alwin; Stamouli, Amalia; Ypma, Rolf JF
• Is Part Of: Forensic science international. Volume 335(2022)
• Publication Date: 2022-06
• Subjects: Comparative GSR analysis -- Likelihood ratio (LR) -- Machine learning -- Evidence evaluation -- Casework data -- Gunshot residue; Medical jurisprudence -- Periodicals; Chemistry, Forensic -- Periodicals; Forensic Medicine -- Periodicals; Médecine légale -- Périodiques; Chimie légale -- Périodiques; Gerechtelijke geneeskunde; Gerechtelijke chemie; Gerechtelijke psychiatrie; Chemistry, Forensic; Medical jurisprudence; Electronic journals; Periodicals; Electronic journals; 614.1
• DOI: 10.1016/j.forsciint.2022.111293
• Languages: English
• ISSNs: 0379-0738

Abstract

Abstract: Comparative gunshot residue analysis addresses relevant forensic questions such as 'did suspect X fire shot Y?'. More formally, it weighs the evidence for hypotheses of the form H1 : gunshot residue particles found on suspect's hands are from the same source as the gunshot residue particles found on the crime scene and H2 : two sets of particles are from different sources . Currently, experts perform this analysis by evaluating the elemental composition of the particles using their knowledge and experience. The aim of this study is to construct a likelihood-ratio (LR) system based on representative data. Such an LR system can support the expert by making the interpretation of the results of electron microscopy analysis more empirically grounded. In this study we chose statistical models from the machine learning literature as candidates to construct this system, as these models have been shown to work well for large and high-dimensional datasets. Using a subsequent calibration step ensured that the system outputs well-calibrated LRs. The system is developed and validated on casework data and an additional validation step is performed on an independent dataset of cartridge data. The results show that the system performs well on both datasets. We discuss future work needed before the method can be implemented in casework. Highlights: Comparative GSR analysis using machine learning methods. A likelihood ratio (LR) system is constructed based on representative casework data. The LR system aims to distinguish the same- and different-source GSR populations. Statistical models from machine learning literature were used to build the system. Gradient boosted trees with logistic calibration for comparative GSR analysis.