Sampling for Forensic Practitioners Short Course

This series is scheduled to begin on Friday, April 28, 2023. A registration form can be found below. 

Presenter:

Alicia Carriquiry
Director of CSAFE
Iowa State University

About the Short Course:

Session 1 is the first session in the three-session short course, Sampling for Forensic Practitioners. Dr. Alicia Carriquiry covers populations and sampling frames, and sampling methods including geometric sampling. The emphasis will be on understanding how these methods are used to aid and enhance current forensic science practices. This short course will be held online in three sessions: April 28, May 5, and May 12. Registration is free, and you only need to register once to attend all three sessions. Participants who attend all sessions will receive a certificate of completion.

If you are interested in participating in this course and would like to suggest topics for discussion, please contact us at your earliest convenience and we will try to accommodate your request.

Statistical Thinking for Forensic Practitioners Short Course

This series is scheduled to begin on Friday, March 31, 2023. A registration form can be found below. 

Session 1 Description:

Probability Concepts and their Relevance to Forensic Science is the first session in the four-session short course, Statistical Thinking for Forensic Practitioners

Probability is the mathematical language of uncertainty. Probabilities are used to describe the frequency or likelihood of events or to characterize measurement uncertainty. In this first session, we introduce the laws of probability and their application in forensic settings. Specific topics include:

  • Definition and interpretation of probability
  • Basic laws of probability
  • Conditional probability and independence of events
  • Bayes’ Theorem and Bayesian statistics

Topics are illustrated with examples drawn from forensic science and relevant legal cases.

Presenter:

Hal Stern
Co-Director of CSAFE
Provost, Executive Vice Chancellor, and Chancellor’s Professor – University of California, Irvine

About the Short Course:

Session 1 is the first session in the four-part short course, Statistical Thinking for Forensic Practitioners, Dr. Hal Stern introduces fundamental concepts from probability and statistics –– motivated by forensic issues –– followed by a detailed investigation of how they apply to assess forensic evidence’s probative value. This short course will be held online in four sessions. Each session builds upon the previous one(s) and recordings will be available in the event registrants are unable to attend one or more of the live sessions. Researchers, collaborators, and members of the broader forensics and statistics communities are encouraged to attend. Short course registrants who attend all sessions will receive a certificate of completion.

Webinar: Judging Firearms Evidence

CSAFE invites researchers, collaborators, and members of the broader forensics and statistics communities to participate in our Spring 2023 Webinar Series on Wednesday, February 22, 2023, from 10:00-11:00am CT. The presentation will be a review of firearms caselaw.

Presenter:
Brandon Garrett
L. Neil Williams, Jr. Professor of Law
Director, Wilson Center for Science and Justice

Presentation Description:

Firearms violence is a major problem in the United States. Each year, over 10,000 homicides involve firearms, and almost 500,000 other crimes, such as robberies and assaults, are committed using firearms. Firearms examiners seek to link fired shell casings or bullets from crime scene evidence to a particular firearm. The underlying assumptions is that firearms impart unique marks on bullets and cartridge cases, and that trained examiners can identify these marks to determine bullets or cartridges cases were fired by the same gun. For over a hundred years, firearms examiners testified in criminal trials that they can conclusively identify the source of a bullet or shell casing. In recent years, however, research scientists have called into question the validity and reliability of such testimony. Further, the revolution in judicial screening of expert testimony following Daubert v. Merrell Dow Pharmaceuticals, Inc., has slowly impacted rulings in criminal cases.

In this presentation, I describe an Article in which we detail over a century of caselaw, examining how judges have engaged with the changing practice and scientific understanding of firearms comparison evidence. We first describe how judges were initially skeptical of firearms comparison evidence and thought that jurors were capable of making the comparisons themselves, without a need for expert testimony. Next, we document how claims made by experts became more specific and aggressive as the work spread nationally. Finally, we explore the modern era of firearms caselaw and research, which has become decidedly more complex. Judges increasingly express skepticism and adopt a range of approaches to limit in-court testimony by firearms examiners. We conclude by examining the lessons regarding the gradual judicial embrace of scientific critiques of expert testimony that can be learned from the more-than-a-century-long arc of judicial review of firearms evidence in the United States.

 

Webinars are free and open to the public, but researchers, collaborators and members of the broader forensics and statistics communities are encouraged to attend. Space is only guaranteed for the first 150 people registered. Each 60-minute webinar will allow for discussion and questions.

Webinar: The ASCLD Forensic Research Committee and You: a Collaboration Worth Investigating

CSAFE invites researchers, collaborators, and members of the broader forensics and statistics communities to participate in our Fall 2022 Webinar Series on Thursday, November 17, 2022, from 11:00am-Noon CT. The presentation will be “The ASCLD Forensic Research Committee and You: a Collaboration Worth Investigating.”

Presenter:
Henry Maynard
Forensic Research Committee Chair, American Society of Crime Laboratory Directors

Presentation Description:

Over the last few years, the American Society of Crime Laboratory Directors (ASCLD) Forensic Research Committee (FRC) has created and launched tools to help advance forensic science research and further research collaborations within the forensic science community. From this presentation, participants will learn about forensic science research needs, the Laboratories and Educators Alliance Program (LEAP) which enables research partnerships, repositories for forensic research, evaluation, and validation efforts, executive research summaries, the research collaboration hub, Lighting Talks, and more! This presentation is for individuals who are looking to become more engaged in forensic science research and want to learn more about the opportunities that the ASCLD FRC is creating to benefit the forensic science community.

 

For more information about these initiatives or other ASCLD FRC Initiatives please review the ASCLD FRC Website: https://www.ascld.org/forensic-research-committee/

 

Webinars are free and open to the public, but researchers, collaborators and members of the broader forensics and statistics communities are encouraged to attend. Space is only guaranteed for the first 150 people registered. Each 60-minute webinar will allow for discussion and questions.

Webinar: handwriter: A Demonstration and Update on CSAFE Handwriting Analysis

CSAFE invites researchers, collaborators, and members of the broader forensics and statistics communities to participate in our Summer 2022 Webinar Series on Tuesday, October 18, 2022, from 11:00am-Noon CT. The presentation will be “handwriter: A Demonstration and Update on CSAFE Handwriting Analysis.”

Presenter:
Alicia Carriquiry
Director, Center for Statistics and Applications in Forensic Evidence (CSAFE)
Distinguished Professor and President’s Chair, Department of Statistics – Iowa State University

Presentation Description:

Forensic handwriting analysis relies on the principle of individuality: no two writers produce identical writing, and given enough quality and quantity of writing, it is possible to infer whether two documents were written by the same person. Forensic handwriting analysis is carried out by examiners trained to find subtle differences and similarities between a questioned document and a reference sample. Examiners visually compare samples and offer an opinion regarding the source of the questioned document. 

In the last 10-15 years, researchers have proposed algorithmic tools to complement examiners’ visual assessments. A well-known software system called FLASH ID (Sciometrics, LLC) first decomposes the image of a questioned sample into structures called graphemes and then characterizes them by their topology and shape. Given a closed set of reference samples, the software computes a score that quantifies the similarity between the questioned document and the references. FLASH ID has been extensively tested and has been shown to exhibit very good accuracy. 

At CSAFE, we are working on semi-automated methods suitable for closed or for open sets of reference writers, and for the examination of samples at the level of words or at the level of graphical structures similar, but not identical to graphemes. In the webinar we will describe each of the different methods, and show initial but promising results. When the reference set of writers is closed we use a Bayesian approach that outputs a probability of writership for each writer in the set. Because the output is an estimated probability, the interpretation of results is straightforward. We are still developing and testing the word-based approach and the more algorithmic approach that can be used when the set of potential writers is open, but can show some initial results and our plans for future developments. We will demonstrate our software to implement these methods: handwriter. The program is not yet fully functional, but an early version is in the public domain and can be freely accessed at https://github.com/CSAFE-ISU/handwriter. 

 

Webinars are free and open to the public, but researchers, collaborators and members of the broader forensics and statistics communities are encouraged to attend. Space is only guaranteed for the first 150 people registered. Each 60-minute webinar will allow for discussion and questions.

Sign up on the form below (Chrome & Safari web browsers work the best):

Webinar: Tutorial on Likelihood Ratios with Applications in Digital Forensics

CSAFE invites researchers, collaborators, and members of the broader forensics and statistics communities to participate in our Summer 2022 Webinar Series on Thursday, September 15, 2022, from 11:00am-Noon CT. The presentation will be “Tutorial on Likelihood Ratios with Applications in Digital Forensics.” 

Presenters:
Rachel Longjohn
PhD Student – University of California, Irvine

Padhraic Smyth
Associate Director – Center for Machine Learning and Intelligent Systems, University of California, Irvine

Presentation Description:

To date, digital forensics research has largely focused on extracting and reconstructing information from devices and the cloud. In comparison, there has been relatively little work on statistical methodologies that can be used to analyze such data after this step. In this webinar, we will discuss statistical analyses in digital forensics, with a particular focus on likelihood ratios and ideas from Bayesian statistics. There will be three parts to the webinar.

First, we will begin with a general introduction to the concept of likelihood ratios. We will show how they can be constructed mathematically, how they can be interpreted, and how they have been broadly applied in forensics. We will discuss how strategies from Bayesian statistics can be incorporated into the statistical models used to construct the likelihood ratio and walk through simple motivating examples step-by-step.

Second, we will discuss the development of likelihood ratios in the context of digital forensics. We will consider the types of evidence available in digital forensics, the types of questions investigators may ask about this data, and how likelihood ratios can be used to address these questions. Building upon the first part of the webinar, we will present a likelihood ratio-based method for analyzing digital evidence data which uses a Bayesian approach.

Lastly, we will present results from applying these methods to real-world datasets related to digital evidence. We will discuss these results, limitations of the method, and how future research can improve upon this approach.

 

The webinars are free and open to the public, but researchers, collaborators and members of the broader forensics and statistics communities are encouraged to attend. Each 60-minute webinar will allow for discussion and questions.

 

Machine Learning for Forensic Practitioners Short Course – Session 2

This event is scheduled to take place on Thursday, September 8, 2022. It is the second session of a three-part short course. A registration form can be found below. 

 

About the Short Course:
The use of learning algorithms will increase as measurement of features in various types of evidence improve. This is particularly true in the case of pattern evidence. Forensic scientists will greatly benefit from understanding the basic ideas that underpin statistical learning since these types of methods have already been proposed for firearms examination, fingerprints, glass comparison, and shoe print evidence. Most quantitative training for forensic scientists emphasize classical statistical ideas, so a workshop in which forensic practitioners are exposed to learning algorithms is novel and timely.
When a task consists of deciding whether two items are similar enough to suggest that they could have a common source, an alternative approach is to use statistical or machine learning. Machine learning is the term used to refer to a family of statistical methods and computer algorithms that find patterns in data and has been around for decades. There are many different types of algorithms, but a basic taxonomy is to distinguish between supervised learning algorithms and unsupervised learning algorithms. The purpose of this short course is on supervised learning methods.
Supervised algorithms rely on training data, for which ground truth is known, and on test data, on which the performance of the algorithm can be tested. In a simple example, several bullets are fired from a large number of guns. To train an algorithm to recognize whether a pair of bullets was fired from the same or from a different gun, one might begin by creating all possible pairs of bullets, and compute, for example, the difference in the average striation depth for each pair. This difference is a feature, and perhaps it takes on low values when two bullets were fired from the same gun and high values otherwise. Presented with the value of the feature for pairs of bullets known to have been fired from the same or from different guns, the algorithm then “learns” that same-gun bullets tend to exhibit values of the feature in a certain range that is different for different-gun bullets. With this knowledge, the algorithm can then classify other pairs of bullets for which it does not know in advance whether the bullets were shot from the same or from a different gun.
In real applications, the number of features can be very large, and the number of classes can also be large. In classification examples, the response variable—or class—is discrete, but algorithms can also be used when the response is continuous; in this case, the problem is to predict the value of a variable given information on a large number of features.
Each session builds upon the previous one(s) and recordings will be available in the event registrants are unable to attend one or more of the live sessions. Researchers, collaborators, and members of the broader forensics and statistics communities are encouraged to attend. Short course registrants who attend all sessions will receive a certificate of completion.

Presenters:
Heike Hofmann
Researcher, CSAFE
Kingland Data Analytics Faculty Fellow and Professor, Iowa State University

Alicia Carriquiry
Director, CSAFE
Distinguished Professor of Liberal Arts and Sciences and Professor of Statistics, Iowa State University

Jeff Salyards
Research Scientist, CSAFE

Registration:

Machine Learning for Forensic Practitioners Short Course – Session 3

This event is scheduled to take place on Thursday, September 15, 2022. It is the final session of a three-part short course. A registration form can be found below. 

 

About the Short Course:
The use of learning algorithms will increase as measurement of features in various types of evidence improve. This is particularly true in the case of pattern evidence. Forensic scientists will greatly benefit from understanding the basic ideas that underpin statistical learning since these types of methods have already been proposed for firearms examination, fingerprints, glass comparison, and shoe print evidence. Most quantitative training for forensic scientists emphasize classical statistical ideas, so a workshop in which forensic practitioners are exposed to learning algorithms is novel and timely.
When a task consists of deciding whether two items are similar enough to suggest that they could have a common source, an alternative approach is to use statistical or machine learning. Machine learning is the term used to refer to a family of statistical methods and computer algorithms that find patterns in data and has been around for decades. There are many different types of algorithms, but a basic taxonomy is to distinguish between supervised learning algorithms and unsupervised learning algorithms. The purpose of this short course is on supervised learning methods.
Supervised algorithms rely on training data, for which ground truth is known, and on test data, on which the performance of the algorithm can be tested. In a simple example, several bullets are fired from a large number of guns. To train an algorithm to recognize whether a pair of bullets was fired from the same or from a different gun, one might begin by creating all possible pairs of bullets, and compute, for example, the difference in the average striation depth for each pair. This difference is a feature, and perhaps it takes on low values when two bullets were fired from the same gun and high values otherwise. Presented with the value of the feature for pairs of bullets known to have been fired from the same or from different guns, the algorithm then “learns” that same-gun bullets tend to exhibit values of the feature in a certain range that is different for different-gun bullets. With this knowledge, the algorithm can then classify other pairs of bullets for which it does not know in advance whether the bullets were shot from the same or from a different gun.
In real applications, the number of features can be very large, and the number of classes can also be large. In classification examples, the response variable—or class—is discrete, but algorithms can also be used when the response is continuous; in this case, the problem is to predict the value of a variable given information on a large number of features.
Each session builds upon the previous one(s) and recordings will be available in the event registrants are unable to attend one or more of the live sessions. Researchers, collaborators, and members of the broader forensics and statistics communities are encouraged to attend. Short course registrants who attend all sessions will receive a certificate of completion.

Presenters:
Heike Hofmann
Researcher, CSAFE
Kingland Data Analytics Faculty Fellow and Professor, Iowa State University

Alicia Carriquiry
Director, CSAFE
Distinguished Professor of Liberal Arts and Sciences and Professor of Statistics, Iowa State University

Jeff Salyards
Research Scientist, CSAFE

Registration:

Machine Learning for Forensic Practitioners Short Course – Session 1

This event is scheduled to take place on Thursday, September 1, 2022. It is the first session of a three-part short course. A registration form can be found below. 

 

About the Short Course:
The use of learning algorithms will increase as measurement of features in various types of evidence improve. This is particularly true in the case of pattern evidence. Forensic scientists will greatly benefit from understanding the basic ideas that underpin statistical learning since these types of methods have already been proposed for firearms examination, fingerprints, glass comparison, and shoe print evidence. Most quantitative training for forensic scientists emphasize classical statistical ideas, so a workshop in which forensic practitioners are exposed to learning algorithms is novel and timely.
When a task consists of deciding whether two items are similar enough to suggest that they could have a common source, an alternative approach is to use statistical or machine learning. Machine learning is the term used to refer to a family of statistical methods and computer algorithms that find patterns in data and has been around for decades. There are many different types of algorithms, but a basic taxonomy is to distinguish between supervised learning algorithms and unsupervised learning algorithms. The purpose of this short course is on supervised learning methods.
Supervised algorithms rely on training data, for which ground truth is known, and on test data, on which the performance of the algorithm can be tested. In a simple example, several bullets are fired from a large number of guns. To train an algorithm to recognize whether a pair of bullets was fired from the same or from a different gun, one might begin by creating all possible pairs of bullets, and compute, for example, the difference in the average striation depth for each pair. This difference is a feature, and perhaps it takes on low values when two bullets were fired from the same gun and high values otherwise. Presented with the value of the feature for pairs of bullets known to have been fired from the same or from different guns, the algorithm then “learns” that same-gun bullets tend to exhibit values of the feature in a certain range that is different for different-gun bullets. With this knowledge, the algorithm can then classify other pairs of bullets for which it does not know in advance whether the bullets were shot from the same or from a different gun.
In real applications, the number of features can be very large, and the number of classes can also be large. In classification examples, the response variable—or class—is discrete, but algorithms can also be used when the response is continuous; in this case, the problem is to predict the value of a variable given information on a large number of features.
Each session builds upon the previous one(s) and recordings will be available in the event registrants are unable to attend one or more of the live sessions. Researchers, collaborators, and members of the broader forensics and statistics communities are encouraged to attend. Short course registrants who attend all sessions will receive a certificate of completion.

Presenters:
Heike Hofmann
Researcher, CSAFE
Kingland Data Analytics Faculty Fellow and Professor, Iowa State University

Alicia Carriquiry
Director, CSAFE
Distinguished Professor of Liberal Arts and Sciences and Professor of Statistics, Iowa State University

Jeff Salyards
Research Scientist, CSAFE

Registration:

Webinar: Ensemble SLRs for Forensic Evidence Comparison

CSAFE invites researchers, collaborators, and members of the broader forensics and statistics communities to participate in our Summer 2022 Webinar Series on Thursday, August 25, 2022, from 11:00am-Noon CT. The presentation will be “Ensemble SLRs for Forensic Evidence Comparison.” 

Presenters:
Danica Ommen
Assistant Professor – Department of Statistics, Iowa State University

Federico Veneri Guarch
Graduate Research Assistant – Department of Statistics, Iowa State University

Presentation Description:

To strengthen the statistical foundations of forensic evidence interpretation, likelihood ratios and Bayes factors are advocated to quantify the value of evidence. Both methods rely on formulating a statistical model, which can be challenging for complex evidence. Machine learning score-based likelihood ratios have been proposed as an alternative in those cases. Under this framework, a (dis)similarity score and its distribution under alternative propositions are estimated using pairwise comparisons, but pairwise comparisons of all the evidential objects result in dependent scores. While machine learning methods may not require distributional assumptions, most assume independence. We introduce a sampling and ensembling approach to remedy this lack of independence. We generate sets where assumptions are met to develop multiple score-based  likelihood ratios later aggregated into a final score to quantify the value of evidence.

 

The webinars are free and open to the public, but researchers, collaborators and members of the broader forensics and statistics communities are encouraged to attend. Each 60-minute webinar will allow for discussion and questions.

Sign up on the form below (Chrome & Safari web browsers work the best):