Published: 8th February 2021 DOI: 10.4204/EPTCS.333 ISSN: 2075-2180 |
Following the inaugural meeting in Leiden, Holland and a successful event in Oxford, UK, the third annual International Applied Category Theory Conference (ACT2020) was planned to take place at MIT in Cambridge, Massachusetts USA. However, the global COVID-19 pandemic made the prospect of holding a large in-person meeting impossible. While many 2020 conferences were being canceled, the local organizing committee—and in particular the technical lead Paolo Perrone—both foresaw the seriousness of the pandemic and the possibility of pivoting to a completely online conference, and was able to reverse course quite seamlessly.
The event was thus held completely online. While this had a rather large effect on our ability to interact and collaborate with one another, the number and quality of contributions was almost identical to previous years. Moreover, holding the talks online had the new benefits of reducing carbon footprint, being inclusive of people from more parts of the world, and producing higher-quality video talks, which have been posted online for posterity.
The ACT2020 contributions spanned a broad spectrum of subjects and application areas, and they ranged from quite pure to quite applied. Probably the most popular mathematical subject this year was that of lenses and their generalizations, which were discussed both on a formal level and in applications to databases, dynamical systems, functional programming, game theory, and neuroscience. Also popular were sheaf- and topos-theoretic techniques, used to model compositionality of behavior in cyberphysical, event-based systems, and even more general systems. Categorical probability and statistics was a major theme, with some new advances in stochastic processes, probabilistic programming, and general theory. There was also new work on natural language processing, categorical quantum mechanics, and the interactions between the two subjects.
Many other topics were represented at the conference, including a categorical interpretation of the work of C.S. Pierce, a new algorithm for learning Boolean circuits, and advances in topological data analysis. Of special interest was work on producing software tools for applied category theory, including both a Python library for modeling monoidal categories and a Julia library for compositional scientific computing.
Papers in this Proceedings volume represents about half of the talks presented at ACT2020. Being included in the proceedings vs. not is not an indication of talk quality, but instead almost exclusively the choice of the authors, e.g. to present work already published elsewhere. We hope that readers of this volume enjoy this representative selection and check out the other talks recorded online (a link can be found at act2020.mit.edu), and we look forward to seeing you at the next event!
David Spivak and Jamie Vicary
Co-chairs of the ACT2020 Scientific Committee