Starting with a bachelor’s degree in evolutionary biology I moved on to a rather unrelated though interesting Honours project. I designed and built a 3D printer that used urealytic bacteria to cement sand grains together creating 3D sandstone prints! While this was successful, my Honours taught me that working in a wet lab wasn’t for me. Deciding data science and working at a computer was more my speed I followed my botanical interests to the remote sensing group to commence a PhD.

My area of study is the classification of plants from hyperspectral imagery, with a focus on overcoming issues caused by intra-specific variation (variation within a species). Essentially, how do you apply a classification model when your training data only covers a small subset of the possible variation seen within a plant species, or when the variation in the training data is large and causes overlap between the spectra of different species. I’m looking into applying recent advances in deep machine learning towards this project, while struggling to avoid being (too) distracted by the wonderful world of all things deep learning.

You can find me on Twitter @_a_hennessy