I am primarily interested in data integration for high dimensional, network data. This can include both methodological work, such as 1., and application-driven work. My ongoing work is of this latter form, and involves applying DIVAS to Human Connectome Project data. I anticipate this work being submitted in this calendar year, but it has also inspired some exciting future methodological work on Jackstraw for DIVAS. As a coauthor on 1., I also help maintain the DIVAS github repository alongside Jack Prothero.

I am also quite interested in the budding intersection between philosophy and statistics. This is something I have termed, moral machine learning, and it is intended to encompass concerns such as algorithmic bias, data privacy, and interpretable methods. Thus far, this has manifest itself as a new course (see STOR 390 in my teaching page) at UNC and a pedagogical preprint reflecting on the preparation and delivery therein. However, I think there is far more to be said, both about best practices in making such a course more mainstream and in theoretical work surrounding the above issues.

Publications

1. Prothero, J.; Jiang, M.; Hannig, J. ; Tran-Dinh Q.; Ackerman A.; Marron, J.S. Data integration via analysis of subspaces (DIVAS). TEST (2024). doi

2. Ackerman, A.; Martin, B.; Tanisha, M.; Edoh, K.; Ward, J.P. High-Dimensional Contact Network Epidemiology. Epidemiologia 2023, 4, 286-297. doi

Preprints

1. Ackerman, A. Moral Machine Learning: Teaching a Course at the Intersection of Applied Statistics and Moral Philosophy (under review at Journal of Statistics and Data Science Education)¹
2. Ackerman, A.; Zhang, Z.; Hannig, J.; Prothero, J.; Marron J.S. Multi-faceted Brain Imaging Data Integration via Analysis of Subspaces (submitted to Psychometrika) arXiv

Back to Homepage