MSMichael E. Sigman

Michael E. Sigman

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Michael E. Sigman has a total of 37 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Machine Learning Subjective Opinions: An Application in Forensic Chemistry

    by Anuradha Akmeemana, Michael E. Sigman

    Simulated data created in silico using a previously reported method were sampled by bootstrapping to generate data sets for training multiple copies of an ensemble learner (i.e., a machine learning (ML) method). The posterior probabilities of class membership obtained by applying the ensemble of ML models to previously unseen validation data were fitted to a beta distribution. The shape parameters for the fitted distribution were used to calculate the subjective opinion of sample membership into one of two mutually exclusive classes. The subjective opinion consists of belief, disbelief and uncertainty masses. A subjective opinion for each validation sample allows identification of high-uncertainty predictions. The projected probabilities of the validation opinions were used to calculate log-likelihood ratio scores and generate receiver operating characteristic (ROC) curves from which an opinion-supported decision can be made. Three very different ML models, linear discriminant analysis (LDA), random forest (RF), and support vector machines (SVM) were applied to the two-state classification problem in the analysis of forensic fire debris samples. For each ML method, a set of 100 ML models was trained on data sets bootstrapped from 60,000 in silico samples. The impact of training data set size on opinion uncertainty and ROC area under the curve (AUC) were studied. The median uncertainty for the validation data was smallest for LDA ML and largest for the SVM ML. The median uncertainty continually decreased as the size of the training data set increased for all ML.The AUC for ROC curves based on projected probabilities was largest for the RF model and smallest for the LDA method. The ROC AUC was statistically unchanged for LDA at training data sets exceeding 200 samples; however, the AUC increased with increasing sample size for the RF and SVM methods. The SVM method, the slowest to train, was limited to a maximum of 20,000 training samples. All three ML methods showed increasing performance when the validation data was limited to higher ignitable liquid contributions. An ensemble of 100 RF ML models, each trained on 60,000 in silico samples, performed the best with a median uncertainty of 1.39x10−2 and ROC AUC of 0.849 for all validation samples.


  • Analyst and machine learning opinions in fire debris analysis

    by Frances A. Whitehead, Mary R. Williams, Michael E. Sigman


  • Open Access

    Promoting transparency in forensic science by integrating categorical and evaluative reporting through decision theory

    by Michael E. Sigman, Mary R. Williams

    Forensic science standards often require the analyst to report in categorical terms. Categorical reporting without reference to the strength of the evidence, or the strength threshold that must be met to sustain or justify the decision, obscures the decision-making process, and allows for inconsistency and bias. Standards that promote reporting in probabilistic terms require the analyst to report the strength of the evidence without offering a conclusive interpretation of the evidence. Probabilistic reporting is often based on a likelihood ratio which depends on calibrated probabilities. While probabilistic reporting may be more objective and less open to bias than categorical reporting, the report can be difficult for a lay jury to interpret. These reporting methods may appear disparate, but the relationship between the two is easily understood and visualized by a simple decision theory construct known as the receiver operating characteristic (ROC) curve. Implementing ROC-facilitated reporting through an expanded proficiency testing regime may provide transparency in categorical reporting and potentially obviate some of the lay jury interpretation issues associated with probabilistic reporting.


  • Open Access

    Convolutional Neural Network Applications in Fire Debris Classification

    by Anuradha Akmeemana, Mary R. Williams, Michael E. Sigman

    Convolutional neural networks (CNNs) are inspired by the visual cortex of the brain. In this work, CNNs, are applied to classify ground truth samples as positive or negative for ignitable liquid residue (ILR+ and ILR−, respectively). Known ground truth samples included laboratory-generated fire debris samples, neat ignitable liquids (ILs), single-substrate (SUB) burned samples and computationally generated (in silico) training samples. The images were generated from the total ion spectra for both training and test datasets by applying a wavelet transformation. The training set consisted of 50,000 in silico-generated fire debris samples. The probabilities generated from the CNN are used to calculate the likelihood ratios. These likelihood ratios were calibrated using logistic regression and the empirical cross-entropy (ECE) plots were used to investigate the calibration of the probabilities of the presence of ILRs (i.e., probability of belonging to class ILR+). The performance of the model was evaluated by the area under the receiver operating characteristic plots (ROC AUC). The ROC AUC for the laboratory-generated fire debris samples and the combined IL and SUB samples was 0.87 and 0.99, respectively. The CNNs trained on in silico data did significantly better predicting the classification of the pure IL (ILR+) and SUB (ILR−) samples. Nonetheless, the classification performance for laboratory-generated samples was sufficient to aid forensic analysts in the classification of casework samples.


  • Open Access

    Detecting Chemical Vapor Diffusion through Firefighter Turnout Gear

    by Michelle A. Corbally, Mary R. Williams, Jessica N. Chappell, Michael E. Sigman

    Firefighters are exposed to burning materials that may release toxic partial combustion and pyrolysis products into the environment, including compounds listed as priority pollutants by the United States Environmental Protection Agency (EPA). A novel passive sampling dosimeter device containing firefighter turnout gear as a diffusion membrane and an activated charcoal strip (ACS) for volatile analyte collection was designed and used to monitor potential exposures of firefighters to volatile organic compounds. Solvent extracts from the ACS and turnout gear diffusion layer were analyzed using Gas Chromatography–Mass Spectrometry (GC-MS) to determine the diffusion of compounds from burned substrates through firefighter turnout gear and compound adsorption to the turnout gear. The compounds in these samples were identified using target factor analysis (TFA). An activated carbon layer (ACL) was added to the dosimeter between the turnout gear and the ACS. The presence of combustion and pyrolysis compounds identified on the ACS in the dosimeter was reduced.


  • Chemometric applications in fire debris analysis

    by Michael E. Sigman, Mary R. Williams

    Chemometric applications related to the analysis of ignitable liquids (ILs) and fire debris samples have evolved over the last 30 years. Most older papers focused on the use of chemometrics for classification of ILs by manufacturer, grade or into classes defined by the American Society for Testing and Material (ASTM) standard test method ASTM E1618. More recently, chemometric studies in fire debris analysis have focused on estimating the strength of the evidence in the form of a likelihood ratio (LR). The progression of chemometric methods from simple classification tasks to the calculation of LRs and statements of evidentiary value has implications across all disciplines of forensic science. Most of the data analyzed by chemometric methods in forensic fire debris analysis come from gas chromatography–mass spectrometry, although several studies have examined data from head space‐mass spectrometry and other sensors. Retention time alignment is a challenge that must be addressed if chemometric methods are intended to have interlaboratory applicability. Several of the reports reviewed here rely on data representations that avoid retention time alignment issues. Attempts to apply chemometric methods to the analysis of ILs and fire debris samples has for the most part been an academic exercise and the results have not made their way into accredited forensic laboratories and established protocols. The chemometric methods are well‐founded in statistical principles and they can provide a means of strengthening forensic fire debris analysis in the future.This article is categorized under: Forensic Chemistry and Trace Evidence > Fire Debris Analysis Forensic Chemistry and Trace Evidence > Trace Evidence Forensic Chemistry and Trace Evidence > Presentation and Evaluation of Forensic Science Output


  • Open Access

    Advances in Fire Debris Analysis

    by Michael E. Sigman, Mary R. Williams

    The practice of forensic fire debris analysis and data interpretation in operational (i


  • Open Access

    Application of Self-Organizing Maps to the Analysis of Ignitable Liquid and Substrate Pyrolysis Samples

    by Nicholas Thurn, Mary R. Williams, Michael E. Sigman

    Classification of un-weathered ignitable liquids is a problem that is currently addressed by visual pattern recognition under the guidelines of Standard Test Method for Ignitable Liquid Residues in Extracts from Fire Debris Samples by Gas Chromatography-Mass Spectrometry, ASTM E1618-14. This standard method does not separately address the identification of substrate pyrolysis patterns. This report details the use of a Kohonen self-organizing map coupled with extracted ion spectra to organize ignitable liquids and substrate pyrolysis samples on a two-dimensional map with groupings that correspond to the ASTM-classifications and separate the substrate pyrolysis samples from the ignitable liquids. The component planes give important information regarding the ions from the extracted ion spectra that contribute to the different classes. Some additional insight is gained into grouping of substrate pyrolysis samples based on the nature of the unburned material as a wood or non-wood material. Further subclassification was not apparent from the self-organizing maps (SOM) results.


  • Open Access

    Model Distribution Effects on Likelihood Ratios in Fire Debris Analysis

    by Alyssa Allen, Mary R. Williams, Nicholas A. Thurn, Michael E. Sigman

    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. The models all perform 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 does 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.


  • Model-effects on likelihood ratios for fire debris analysis

    by Richard Coulson, Mary R. Williams, Alyssa Allen, Anuradha Akmeemana, Liqiang Ni, Michael E. Sigman