Computational Neuroscience (PG)

Module aims

This module provides an introduction to Computational Neuroscience. Students will be provided with an appreciation of the role of computational and theoretical approaches to understanding the nervous system, and practical skills, such as basics knowledge of MATLAB, of numerical tools to simulate models.

Learning outcomes

Understanding key concepts concerning the electrical properties underlying information processing in the nervous system, how networks of individual neurons can together perform computations, and how systems of such networks operate together to solve tasks such as visual perception and motor behaviour; To understand how recorded physiological signals can help us to understand how these systems work, and how engineering and statistical approaches help in the analysis of such data; Appreciation of the current level of development of the field of Computational Neuroscience; Appreciation of the relationship between behavioural control and underlying neuronal principles; Levels of description in modelling in neuroscience; Derivation of algorithms and quantities used in theoretical neuroscience; Basic knowledge of neuroscience simulation in MATLAB; Writing code to simulate basic computational models; Problem solving and Organisational acquired during the three course works.

Module syllabus

Definition and Scope: What is Computational Neuroscience? Different approaches to modelling. Neuron models. Electrical properties of neurons: the cell membrane, integrate-and-fire and conductance-based neuron models, spike rate adaptation, refractoriness, phase-plane analysis of neuron model dynamics, the Hodgkin & Huxley model. Signal Integration: synaptic integration, neuron computation, synaptic physiology, synaptic transmission (channels), synaptic models. Feedforward and Recurrent networks: the perceptron, multilayer networks, balanced network, ring model, rate-based network, spike-based network. Learning and Memory: synaptic plasticity (unsupervised (rate-based, LTP/LTD, STDP) in primary sensory area, supervised learning in cerebellum, reinforcement learning in basal ganglia).

Pre-requisites

Basic linear algebra: vector operations (vector sum, scalar product etc.), matrix and matrix vector multiplication, matrix inversion Basic calculus: differentiation, integration, differential equations, stability analysis, phase plane analysis, bifurcation analysis, basic dynamical systems. Basics of probability theory: random variables, probabilities, probability distributions, joint, marginal and conditional probabilities, Gaussian distribution, stochastic processes Basics of matlab programming: variables, conditional expressions (e.g. if-then-else), loops (e.g. for, while), functions, basic mathematical expressions

Teaching methods

Students will be taught over one term using a combination of lectures, practical lab sessions and study groups. Lecture sessions will be made available on Panopto for review and supplemented with technologies where appropriate to promote active engagement during the lecture such as 'learning catalytics'. Lab sessions and study groups will be based on taught content from lectures to reinforce these topics and allow students to develop their practical lab skills and general understanding.

Lectures: 19 hours
Study groups: 9 hours

Assessments

Overall performance against all LOs in the module will be assessed by a final exam in the summer term.

Examinations:
●  Written exam: 100%% weighting

Rubrics: 1 hour exam, 3 questions, all mandatory (partially multiple choice questions). Questions need to be answered on a answer sheet.
No type of previous exam answers or solutions will be available

Feedback : feedback is given verbally by the GTAs during the 9h sessions.