Caleb Charpentier

I'm Caleb Charpentier, a Ph.D. student in Uyeda Lab at Virginia Tech. I'm interested in the evolution of morphology, and how we can incorporate Knowledge-Guided Machine Learning (KGML) and Imageomics techniques to rethink the ways we study phenomes from raw and high-dimensional datasets.

My research is mostly two-fold: I'm interested in developing deep-learning models that learn structured latent spaces for phenotypic data, and in developing theory and techniques that can better ground these models in the traditional scientific process.

Traditional character construction methods are often inexhaustive, redundant, or difficult-to-reproduce. Deep learning techniques can address these limitations, but many of them are "black-boxes", whose decisions are difficult to interpret, greatly limiting their use in evolutionary biology research. My hope is to resolve these issues, and pave the way for researchers to not simply use machine learning to measure large numbers of traits, but also to change the way we understand traits on an empirical, fundamental level.