Machine learning with digitized specimens
Improvements in genome sequencing technology and the availability of digitized environmental datasets have facilitated using big data approaches to investigate genomic and environmental variation within and among populations and species, but tools to collect high-throughput multi-dimensional phenotypic data have lagged behind.
I have designed workshops that help learners use digitized biological image data in their own research. These workshops aim to get learners past the initial steep learning curve sometimes associated with machine learning applications.I have used machine learning with digitized biological image data to address evolutionary questions in some of my research. Here is an example of using machine learning to test phylogenetically-framed hypotheses.