Particular interests of the Computational Neuroscience unit include:
- Dynamic causal modelling
- Active inference
- Computational psychiatry
- Animal and artificial perception
- Computational modelling in the cerebellum
- Information theory and neuronal data
- Network and nonlinear dynamics
- Neuroimaging data analysis
- Synaptic plasticity
- Statistical methods for neural population data
Working in this area:
We are keen to supervize good students motivated to do PhD research; below is a list of potential areas.
Rui Ponte Costa's group does research on neural and machine learning.
This group brings together computational neuroscience and machine learning. We have the following PhD projects available:
- From machine to neuronal recurrent learning in deep RNNs.
- Deep cortical reinforcement learning.
- What are the neuronal basis of fast relearning?
- Dynamic cortical connectivity.
- StatEI nets: networks of statistical excitation-inhibition balance.
Conor Houghton's group does research on cerebellum, decision making and neurolinguistics.
At the moment he is most interested in supervizing research projects on
- modelling neuromodulation in the cerebellar cortex
- investigating neuronal corellates of language comprehension and learning
- bin-free estimates of information theory quantities for neuronal data.
Nathan Lepora's group does research on biometic computing.
He would consider PhD students in Computational/Mathematical Neuroscience on topics relating to my Leverhulme Leadership award on ‘A biomimetic forebrain for robot touch’ (www.lepora.com/leverhulme), specifically
- System-level models of the forebrain based on statistical optimality
- Related models of perceptual decision making, action selection and active perception
- Embodiment of these models on robots (3D-printed tactile robot hands and tactile whisker arrays)
Naoki Masuda's group does research on networks and neuroscience.
He would consider PhD students in the following topics in neuroscience:
- Network neuroscience: network analysis (also known as graph theory) of MRI, EEG and other
- Neuroimaging data analysis inspired by statistical-physics methods (e.g. Ising model)
- Computational studies of whole-brain dynamics, related to the rest, ageing and diseases (these context also apply to the first two topics)
Cian O’Donnell’s group does research on:
biophysical modelling of synaptic plasticity; neural circuit dynamics in autism; statistical methods for neural population data.
Example projects for each respective topic include:
- Understanding the computational benefits of low-copy number molecular signalling at synapses for learning.
- Analysing and modelling neural responses to sensory stimuli in mouse models of autism, in collaboration with experimental colleagues.
- Developing new machine learning/statistical methods specifically for neural data that is recorded in multiple brain regions simultaneously.