Induction: Analogy, Learning, and Generalisation in Humans and Machines

Description

In this course we will cover topics in human and machine inductive inference. In the first half of the course, students will be exposed to the problem of induction and how the problem manifests in a range of domains such as object recognition, categorisation, and learning. The focus of the course will then turn to analogy and relational reasoning, areas were humans make generalisations across situations and domains with much more success and flexibility than non-human animals and conventional machine learning approaches. We will cover research in analogical reasoning as well as the development of analogical thinking and the representations that support analogy and generalisation. The second half of the course will focus on computational theories of how humans and artificial (i.e., machine) systems perform induction and generalisation. We will cover broadly the main approaches to representing knowledge and modelling human cognition (symbolic and connectionist models). We will then cover how these approaches have been leveraged to explain human induction and learning with focus on traditional production system models, Bayesian models, neural network models, and symbolic-connectionist models. Assessment Information: Written Exam 0%, Coursework 100%, Practical Exam 0%