Scientist and software engineer. PhD in Computational Neuroscience. Based in the Vogels Lab at Oxford University. Main interests: olfactory learning, insect learning, mushroom body, dopamine modulated plasticity.
Modelling olfactory learning using SuperSpike
Recent theoretical results has shown that backpropagation does not need symmetric weights to achieve high performance for training an SNN.
Zenke et al. 2016 has proposed a three-factor learning rule that can learn precise output spike trains called SuperSpike. I implemented a spiking
neural network based on the olfactory circuit of the fly and I adapted the SuperSpike learning rule to the problem of olfactory reinforcement learning.
Modelling dopamine modulated learning in the olfactory circuit of the fruit fly
We developed a spiking neural network based on the known wiring diagram of the olfactory circuit of the fruit fly. We tested a family of learning rules to find a mechanism for learning, forgetting and re-learning that is in accordance with experimental evidence.
I developed Microsoft Kinect video games based on gesture recognition control to study opportunities to use gesture recognition in physical rehabilitation.
University of Oxford
Oct 2014 - January 2019
DPhil in Computational Neuroscience
My DPhil was foncused on modelling how olfactory reinforcement learning is encoded in the fruit fly, Drosophila Melanogaster. I developed a spiking neural network and implemented a series of learning rules
to test how dopamine and plasticity interact. I have also been a teaching assistant for introduction to programming, Graph Theory and Computational neuroscience modules.