Editor’s Letter—Vol. 33, No. 2
Dear CHANCE Colleagues,
It is my hope that you remain healthy and safe during these difficult times created by the COVID-19 pandemic. I’m excited to share the April issue with you as a brief, social-distancing-friendly distraction.
In 1999, CHANCE (Vol. 12, No. 2) included an article analyzing the writings of American folk hero and politician Davy Crockett, considering whether he actually wrote the three books attributed to him. Four years later, in 2003, a special issue of CHANCE (Vol. 16, No. 2) focused on stylometry—the statistical analysis of literary style. In this 2020 issue, David Holmes, a guest editor of the stylometry special issue, and Ferris Samara, a George Mason University OSCAR scholarship recipient, take a fresh look at the writings of Davy Crockett using the latest advances in stylometry. They consider Crockett’s ghostwriters and share their discoveries in “Was the Wild Frontiersman a Prolific Penman? A Stylometric Investigation into the Works of Davy Crockett.”
Many in the CHANCE community have researched, built, and applied machine learning models to exciting applications, but how many of us have considered whether the models are vulnerable to attack? Attacks include poisoning of training data, evading model prediction, exploiting model access to replicate functionality (e.g., model-stealing), and extracting data. As machine learning continues to advance, model security is a concern for both the public and private sectors. In “Deceiving Machines: Sabotaging Machine Learning,” David Trott explains adversarial machine learning and the challenges practitioners face in combating a range of attacks.
CHANCE editors had an opportunity to chat with Cynthia Rudin, a professor at Duke University and advocate of interpretable machine learning. In the interview, she shares her story of getting into the field, aspects of her current work in this area, and a view of what lies ahead in the next decade of statistics and machine learning. We enjoyed her insights and trust you will enjoy them as well.
I’d like to welcome a new contributor to the CHANCE team: Maria Tackett, an assistant professor at Duke University, joins Mine Çetinkaya-Rundel as column co-editor of “Taking a Chance in the Classroom.” In this issue, they discuss the iterative process of data visualization, providing examples of “from drab to fab” class activities.
In other columns, Mary Gray discusses challenges faced by expert statistical witnesses attempting to make their work understandable to the courts in “The Odds of Justice”; Andrew Gelman and Alexey Guzey explain how prominent researchers can get away with misrepresenting data in “Ethics and Statistics”; and Justin Jacobs looks at whether “Three is Greater than Two” in his new column about statistics and sports, “Beyond the Box Score.”
Hope to see you at the Joint Statistics Meetings this summer!
Amanda Peterson-Plunkett