COMP40260 Connectionism and Dynamical Systems

Academic Year 2020/2021

The theory and practice of modeling with artificial neural networks will be presented as it is relevant to the study of human behaviour and cognition. Models are closely related to human cognitive processes in general and to developmental processes and learning in particular. This module does not aspire to teaching contemporary machine learning or deep learning, but rather to examine how connectionist and dynamical models have been and are used in theorising about humans. We will also cover basic concepts from dynamical systems theory and see how these are applied in modelling human behavior.

Show/hide contentOpenClose All

Curricular information is subject to change

Learning Outcomes:

On completing this module, students will have acquired the following knowledge: 1) Understanding of the theoretical and historical foundations of connectionism and artificial neural networks; 2) Understanding of the problem area in which neural networks have usefully been applied to the study of humans, with special focus on work in the 1980's and 1990's; 3) Understanding of the opportunities and limitations of connectionist simulation of human cognitive abilities, with a focus on human development, and, 4) Understanding of the relationship between models and data with specific focus on connectionist models, and 5) learn the basic concepts underlying dynamical systems theory, especially as these have been applied in modeling human behaviour, and will be able to do the following: 5) Design and apply simple neural networks using a module-specific modelling platform (BasicProp), 6) Analyze the performance of a network during and after training, and 7) Relate network performance to the specific details of an empirical problem, and 8) understand the ways in which concepts from dynamical systems theory have been employed in describing human behaviour.

Indicative Module Content:

History and origins of connectionist research; Fundamentals of network architecture and training; Relation of training and testing data to both model performance and to assumptions of researchers; Application of network modelling to human development and learning; Basic concepts of dynamical systems theory; Worked examples of the application of dynamical models to human behaviour.

Student Effort Hours: 
Student Effort Type Hours
Lectures

24

Computer Aided Lab

14

Specified Learning Activities

14

Autonomous Student Learning

88

Total

140

Approaches to Teaching and Learning:
Lectures are provided in the form of short videos available through the BrightSpace Platform. Mandatory and optional readings are also provided through Brightspace. Labs will be conducted with group online classes, using the BasicProp software platform. 
Requirements, Exclusions and Recommendations
Learning Recommendations:

Familiarity with elementary statistics, including distributions and the theories of linear correlation and regression will be assumed.


Module Requisites and Incompatibles
Not applicable to this module.
 
Assessment Strategy  
Description Timing Open Book Exam Component Scale Must Pass Component % of Final Grade
Essay: Brief essay (2000 words) Coursework (End of Trimester) n/a Graded No

25

Continuous Assessment: Three equally weighted small written exercises Throughout the Trimester n/a Graded No

75


Carry forward of passed components
No
 
Resit In Terminal Exam
Autumn No
Please see Student Jargon Buster for more information about remediation types and timing. 
Feedback Strategy/Strategies

• Feedback individually to students, post-assessment

How will my Feedback be Delivered?

Written feedback on all assignments and essay will be provided in a timely manner.

Timetabling information is displayed only for guidance purposes, relates to the current Academic Year only and is subject to change.

 
Spring
     
Lecture Offering 1 Week(s) - 18, 19, 20, 21, 22, 23, 24, 27, 28, 29, 30 Mon 11:00 - 12:50
Practical Offering 1 Week(s) - 18, 19, 20, 21, 22, 23, 24, 27, 28, 29, 30 Mon 14:00 - 15:50
Spring