SCIENTIFIC COMPUTING AND IMAGING INSTITUTE
at the University of Utah

An internationally recognized leader in visualization, scientific computing, and image analysis

Kenneth Blake Vernon - One-U Responsible AI Postdoctoral Fellow

This email address is being protected from spambots. You need JavaScript enabled to view it.
supervisor Simon C. Brewer and Brian F. Codding
Personal Home Page

Background

  • University of Utah | Salt Lake City, UT
    PhD in Anthropology, 2022
  • Northern Illinois University | DeKalb, IL
    MA in Philosophy, 2009
  • University of Central Arkansas | Conway, AR
    BA in History and Philosophy, 2006

Current Responsibilities

My work as a One-U Responsible AI Postdoctoral Fellow is narrowly focused on a specific aspect of my broader research, namely, climate adaptation in agricultural systems, particularly those related to water usage. The immediate goal is to develop an encoder-decoder hydrological model (EDHM) that can estimate soil moisture and streamflow, both of which have obvious importance to farmers. The larger ambition, though, is to contribute to the overall mission of One-U RAI - to the promotion of responsible AI - by making that modeling effort as open, as interpretable, and as accessible as possible to both scientists and community stakeholders. The hope is that this will contribute to strategic climate adaptation planning, too.

Research Interests

Broadly speaking, my research focuses on the ecology of human behavior, with a particular emphasis on spatial dynamics. Here, though, is a list of the specific questions that really animate me these days:
  • Climate adaptation and migration: in very abstract terms, how do people distribute themselves across space, and how is that influenced over the long term by climate trends? For this, I usually draw on species distribution modeling approaches from ecology, including machine learning approaches like Random Forest and Maximum Entropy, though I also find my attention drawn more and more to INLA models, too.
  • The micro-economics of early, early, early urbanism: like, how do we get from small, dispersed groups of mobile hunter-gatherers to large concentrations of sedentary city dwellers? I'm attracted to this problem mostly because it's a spatial problem, but there's also this other really weird thing about cities that I find deeply fascinating: how it is that they can exhibit such dramatic scaling effects while still failing utterly to resolve the problem of collective action. 
  • Demographic interpolation: how do we estimate population size without a census? There is a lot of neat work on this using remote-sensed measures of built area to interpolate modern populations across a grid, but transferring those efforts over to deep time raises an interesting problem, namely, that built area in that context is a palimpsest (like the many layers of ancient Troy), which requires that we estimate both a chronology and a population size at the same time. The current statistical approaches I find interesting for this include Gaussian and Dirichlet Process Mixture Models, as well as their deep learning analogs.
In my spare time, I contribute to several FOSS projects in the R community.