Andrew King - Neuroscience - 'Models of the encoding of sounds by the auditory system'
Neurons in auditory cortex produce patterns of spikes in response to sound. What is the computational mapping from sound to auditory cortical responses? Can we predict the moment-to-moment responses of neurons in the auditory cortex to arbitrary sounds? This computational project involves fitting various models to the mapping from natural sounds to neural responses recorded from the midbrain and auditory cortex. The project will begin with a fitting a simple model, the linear-nonlinear model, and testing its capacity to predict neural responses to a held-out dataset of sounds. The student will then go on to analyse the linear and nonlinear parts of this model to look for different types of neuron and extend this work to include more complex network models that better reflect the physiological properties of auditory neurons and the way information processing changes between the midbrain and cortex. The structure of these more complex networks will be analysed to provide insight into the functional organization of the auditory brain.
Other computational and experimental projects on auditory and multisensory processing will also be available.