Michael is interested in the emergent unexpected behaviours of neuroscientific models, focusing on dendritic nonlinearities. In his work, he studies how interactions between biological mechanisms can create phenomena that exceed those expected from each mechanism individually, impacting computation and learning in the brain in unexpected and informative ways.

Using single neuron modelling and dynamic system analysis, Michael studies the phenomena found in the interactions between the spatio-temporal input space of neurons and the ion channels covering the dendritic membrane.

Using probabilistic programming and data science algorithms, Michael works on developing tools for automatically highlighting those interesting phenomena, thus facilitating scientific discovery in computational models.