Application of Chemometrics and Fast GC-MS Analysis for the Identification of Ignitable Liquids in Fire Debris Samples
Author | : Michael E. Sigman |
Publisher | : |
Total Pages | : 0 |
Release | : 2012 |
ISBN-10 | : OCLC:826637658 |
ISBN-13 | : |
Rating | : 4/5 (58 Downloads) |
Download or read book Application of Chemometrics and Fast GC-MS Analysis for the Identification of Ignitable Liquids in Fire Debris Samples written by Michael E. Sigman and published by . This book was released on 2012 with total page 0 pages. Available in PDF, EPUB and Kindle. Book excerpt: The goal of the research conducted under this grant was to develop a chemometric method of data analysis that would facilitate the identification of GC-MS patterns associated with ignitable liquid classes, as designated under ASTM E 1618-10. The objective of the research was to develop a data analysis method that would classify ignitable liquid residue in the presence of background interferences found in fire debris. Pattern recognition and classification methods available at the onset of this research did not explicitly take into account background interference issues. A novel method was developed under this research to classify ignitable liquid residues into the ASTM classes, even in the presence of a strong background signal, without a priori knowledge of the background signature. The method makes use of target factor analysis (TFA) in combination with Bayesian decision theory. The use of Bayesian decision theory provides results in the form of posterior probabilities that a set of samples from a fire scene contain an ignitable liquid of a specific ASTM class. Error rates are not currently available for fire debris analysis, other than extrapolations from proficiency tests. The method was further refined by introducing a sensitivity parameter which made the method very conservative in its predictions, and gave a true "soft" classifier. Soft classifiers allow classification of a sample into multiple classes and afford the possibility of not assigning the sample to any of the available classes. In order to achieve the goals, this work was broken down into three tasks.