Using machine learning algorithms for chemical and feature comparison tasks
We are creating curated ground truth (known source) data sets and applying machine learning algorithms to explore the interpretation of chemical profiles and feature comparison problems.
What we are doing
We are exploring the use of artificial intelligence and machine learning to explore how we recognise patterns in data, both from instruments (such as drug analysis or DNA profiles) and feature comparisons (such as ballistics, fingerprints, toolmarks and shoe prints).
Why we are doing it
The purpose of this work is to explore how we can use AI and machine learning tools to address the subjective nature of comparison of questioned and reference data by creating objective mathematical ways to triage data. These can be used to inform the interpretation and evaluation of evidence encountered in case work.
How we will do it
We are initially working of ignitable liquid data to explore the ability of algorithms to compare linkages and differentiation of samples.