The University of Sheffield
Department of Computer Science

COM3524 Bioinspired Computing

Summary

This module focuses on modern artificial intelligence (AI) techniques and their inspiration from biological systems. Examples include evolution, multicellular tissues, neural systems, the immune system and swarms, inspiring abstractions such as evolutionary or swarm-based optimization algorithms, neural computing, as well computational approaches to simulate real world systems, (e.g. cellular automata and agent-based models). Lecture material introduces a range of AI and related approaches in the context of their relevant biological inspiration and also their potential application to real word problems. A selection of optimisation and simulation techniques are explored in more depth using Python via active learning exercises. There is an emphasis on applying the scientific approach to practical work within this module.

Session Autumn 2021/22
Credits 10 credits
Assessment
  • Group Project
  • Formal Exam
Lecturer(s) Dr Dawn Walker & Dr Paul Watton
Resources
Aims

This module aims to:

  • provide a foundation in modern computational and AI techniques inspired by specific features of biological systems;
  • provide experience of collaborative work that develops biologically-inspired solutions to practical problems;
  • provide experience of using the scientific method and critical analysis skills to explore a particular question or hypothesis in the context a bioinspired method or algorithm
Objectives

On completion of this module, students will be able to:

  • describe essential features of biological inspired AI
  • explain the key biological features and concepts that have inspired different approaches
  • select the most appropriate bioinspired algorithm for a particular purpose
  • evaluate the accuracy and efficiency of different bio-inspired optimisation approaches to solve a real world problem
  • apply bio-inspired computing techniques to investigate a real world problem
Content Indicative content include evolutionary computing, cellular-inspired computing, swarm-based systems and neural-inspired systems (full details to be confirmed).
Restrictions Available to students in Computer Science only. Students must have existing coding skills in Python. Not available to students who have taken COM2005.
Teaching Method Lecture material conveys the key concepts of biological systems and bio-inspired AI approaches. Practical exercises allow the students to apply and explore the concepts that have been covered in lectures. These activities will also work towards formative and summative assessments focussed on e.g. comparing and evaluating two different optimisation algorithms for solving a real world problem or applying a bio-inspired modelling methodology to understand and simulate a real world system.
Feedback Feedback will be provided via mock Blackboard tests or through the use of related interactive questions in live sessions. In addition, students will have the opportunity to be provided with feedback on their work during these sessions by teaching staff/demonstrators.
Recommended Reading
Bio-Inspired Artificial Intelligence - Theories, Methods, and Technologies By Dario Floreano and Claudio Mattiussi https://mitpress.mit.edu/books/bio-inspired-artificial-intelligence