Unit information: Computational Neuroscience in 2011/12

Unit name	Computational Neuroscience
Unit code	COMS30127
Credit points	10
Level of study	H/6
Teaching block(s)	Teaching Block 2 (weeks 13 - 24)
Unit director	Dr. Bogacz
Open unit status	Not open
Pre-requisites	Students from Psychology only have to take “Cognitive modelling in psychology”, PSYC 31039
Co-requisites	None
School/department	Department of Computer Science
Faculty	Faculty of Engineering

Description including Unit Aims

This unit introduces the quantitative theory and models of computations performed by the brain. The lectures will progress through different level of abstraction: from detailed models of single neuron based on neurophysiology to high-level models of interactions between networks of neurons explaining human behaviour. Computational neuroscience is an interdisciplinary subject hence this unit is addressed to students of various backgrounds: computer science, psychology and engineering mathematics.

The following topics are discussed:

• Models of single neuron: integrate and fire model based on neurophysiology.
• Models of neural networks: models of different types of memory in hippocampus and cortex, models of feature extraction in thalamus and visual cortex.
• Models of neural systems: models of decision making, reinforcement learning and cognitive control.

Intended Learning Outcomes

After successful completion of this unit, the student will

• Be inspired by the computational principles of the brain in their future engineering work.
• Be prepared to do research on the brain with understanding of brain’s purpose (i.e., information processing).
• For each levels of abstraction (single neuron, network of neurons, interacting brain areas): understand the assumptions made by the models, validity of the assumptions, and computational principles.
• Be able to simulate simple models of neurons, networks, and cortical areas in Matlab.

Teaching Information

20 hours of lectures, 10 hours of laboratory sessions

Assessment Information

coursework (20%), exam (80%)

All students need to do two courseworks in Matlab which address the 4th learning outcome (Be able to simulate simple models of neurons, networks, and cortical areas in Matlab).

Due to initial differences in background of students attending the unit, the first coursework differs between the students from the Faculty of Engineering and School of Experimental Psychology. In particular, the students from Faculty of Engineering have to simulate activity of two mutually connected neurons, while the students from the School of Experimental Psychology (who learn Matlab as a part of the unit) have to program a psychological experiment of their choice in Matlab. The second coursework is the same for all students and involves simulating neuronal activity during decision making.

Reading and References

Lecture notes. Background reading to include:

• Dayan P & Abbott LF (2001) Theoretical Neuroscience: Computational and Mathematical Modelling of Neural Systems, MIT Press.
• O’Reilly RC & Munataka Y (2000) Computational Explorations in Cognitive Neuroscience, MIT Press.
• Feng J (Ed.) (2003) Computational Neuroscience: A Comprehensive Approach, Chapman & Hall.
• Eliasmith C & Anderson C (2002) Neural Engineering: Computation, Representation and Dynamics in Neurobiological Systems, Bradford Book.
• Rolls E & Deco G (2002) Computational Neuroscience of Vision, Oxford University Press.