The University of Birmingham
School of Computer Science

MSc/PGDip/PGCert in Natural Computation


Introduction

Natural computation is the study of computational systems that use ideas and get inspirations from natural systems, including biological, ecological and physical systems. It is an emerging interdisciplinary area in which techniques and methods are studied for dealing with large, complex, and dynamic problems. This exciting programme will cover a number of topics, such as evolutionary algorithms, co-evolution, evolutionary design, nature-inspired optimisation techniques, evolutionary games, novel learning algorithms, artificial neural networks, theory of natural computation, molecular computation and quantum computation. The primary aim of this advanced MSc is to provide a solid foundation in natural computation for graduates to pursue a research and/or development career in industry or to pursue further studies (e.g., PhD).


Entry Requirements

A very good honours degree or equivalent in Computer Science/Engineering or a closely related area.


Industrial Advisory Board

This MSc programme is supported by the EPSRC through its Master's Level Training Packages and by a number of leading companies. An Industrial Advisory Board has been set up to provide advices and feedback to the MSc programme. The current Board members include:


Course Structure

This is an advanced 12 month programme consisting of 180 credits. The whole programme is divided into three terms of 60 credits each. Students who do not wish to pursue the whole MSc programme may enroll in Postgraduate Certificate (PGCert, 60 credits) or Postgraduate Diploma (PGDip, 120 credits). Part-time students are welcome.

The programme has a strong emphasis on research and research skills in comparison with some other taught MSc programmes. The following paragraphs give an overview of this advanced MSc programme. More up-to-date module descriptions and syllabuses can be found at School's degree regulation page.

Term I

There are three taught modules (10 credits each) and one mini-project (30 credits). All are compulsory.

  1. Introduction to Molecular and Quantum Computation (10 credits): This module covers the basic ideas, techniques and algorithms in molecular and quantum computation. Examples of such ideas, techniques and algorithms include DNA algorithms, molecular gates, parallel computation models executable by DNA-based chemical processes, DNA inference engine, quantum Turing machine, quantum cryptography, quantum search, evolving quantum algorithms, and computation with executable media.

  2. Introduction to Neural Computation (10 credits): This module provides the first introduction to the most common topics in neural computation, including biological neural networks, artificial neuron models, feed-forward artificial neural networks, learning and generalisation, back-propagation, radial basis function networks, self-organising maps, and committee machines.

  3. Introduction to Evolutionary Computation (10 credits): Evolutionary computation is the study of computational systems that use ideas and get inspirations from natural evolution. Its techniques can be applied to optimisation, learning and design. Example topics covered in this module include natural and artificial evolution, genetic algorithms, chromosome representations, search operators, genetic programming, automatically defined functions, schema theorems, and classifier systems.

  4. Mini-project (30 credits): This consists of two parts. The first is a taught module on Research Skills and the second is a research training project that aims at providing various transferable skills that are necessary for conducting successful research, includes literature search, critical literature review, oral and written communication skills, scientific paper and report writing, problem solving, etc.

Term II

There are three taught advanced modules (10 credits each) and one mini-project (30 credits). All are compulsory. The three advanced modules are organised around problems (i.e., optimisation, learning and design), not techniques.

  1. Nature Inspired Optimisation (10 credits): This module introduces a range of nature-inspired techniques and algorithms for both real-valued and combinatorial optimisation. Examples of such techniques and algorithms include global optimisation by evolutionary algorithms, self-adaptation, advanced operators, ant colony optimisation, swarm optimisation, simulated annealing, hybrid algorithms, constrained handling techniques, multi-objective optimisation, optimisation in a dynamic environment.

  2. Nature Inspired Learning (10 credits): This module presents main learning techniques inspired by natural, physical and biological systems. Examples of such techniques include learning classifier systems, competitive and cooperative coevolution, evolutionary games, learning through genetic programming, support vector machines, adaptive resonance theory neural networks, Boltzman neural networks, interaction between learning and evolution, Red Queen effect, Baldwin effect, Lamarkian evolution, memetic algorithms.

  3. Nature Inspired Design (10 credits): This module introduces the basic ideas of nature-inspired design techniques. Different algorithms and their applications will be presented. Similarities and differences between these techniques/algorithms and other classical techniques will be discussed whenever appropriate. The design domain ranges from architectural, engineering and graphics design to electronic circuit design and evolvable hardware. Examples of such techniques include evolution and knowledge discovery, circuit design by evolution, novel architectural design by evolution, creative design, interactive evolution, evolutionary graphics, knowledge extraction from evolution, extrinsic evolvable hardware, intrinsic evolvable hardware, on-line adaptation, and implementation issues.

  4. Mini-project (30 credits): This module consists of a research project on the in-depth investigation of a chosen topic coming from industry (strongly encouraged) or academe. The topic must be different from that chosen for the first mini-project in Term I because it is crucial for a student to gain in-depth knowledge in different areas and to be able to apply natural computation techniques to different problems. This mini-project is expected to be extended further into the summer project in the following term.

Term III (Summer Term)

There is a compulsory project (60 credits) in the summer, solving a substantial real world problem using natural computation techniques (including hybrid techniques). Industrial co-supervisors will be used whenever appropriate. This project requires students to apply the knowledge and skills they acquired previously in the course to solve a difficult real-world problem.


Some Projects Proposed by Industry


Tuition Fees

In 2001/02, these were GBP2805 per year for UK/European Union students; GBP9600 per year for overseas students. More details can be found from the University webpages.

Financial Support

  1. Several fully funded EPSRC studentships, covering fees and maintenence costs, are available each year. They will be awarded on a competitive basis according to applicants' academic achievements. We are sorry that international students are not eligible for these studentships.

  2. In addition to the above funding which is specific to this course, students (both UK and international) are also eligible to apply for many other studentships and scholarships.

How to Apply


Contact for Further Information

The Admissions Secretary
   School of Computer Science
   University of Birmingham
   Edgbaston
   Birmingham
   B15 2TT
   United Kingdom

   Within the UK:
       phone: 0121-414 4782
       fax:   0121-414 4281

   Outside the UK:
       phone: +44 121-414 4782
       fax:   +44 121-414 4281

   Email:  Admissions@cs.bham.ac.uk

   Admissions Tutor and Course Director: Prof. Xin Yao. Email: x.yao@cs.bham.ac.uk

Computer Science Home Page (http://www.cs.bham.ac.uk/)

+++++++++++++++++++++++++
Page maintained by: Prof. Xin Yao. Email: x.yao@cs.bham.ac.uk.