University Homepage School Homepage School Contacts School Search

GAGP Lecture Slides

• Lecture 1. A simple genetic algorithm example: allocating tutors to tutorials.
  • Lecture 1 1up.
  • Lecture 1 4up.
• Lecture 2. Genetics and evolution. Biological background to GAs.
  • Lecture 2 1up.
  • Lecture 2 4up.
  • DNA molecule diagram from Access Excellence Graphics Gallery.
  • Relationship between genotype and phenotype from Blamire's Science at a Distance.
  • Reading Mitchell pages 1 to 27.
• Lecture 3. The canonical GA.
  • Lecture 3 1up.
  • Lecture 3 4up.
  • Reading Mitchell pp. 1-27 if you've not read this already.
  • Reading From chapter 2 of lecture notes: read Whitley's GA tutorial sections 1 and 2.
• Lecture 4. Evolving Strategies
  • Lecture 4 1up.
  • Lecture 4 4up.
  • Refer to lecture notes section 6.4.
• Lecture 5. The Schema Theorem
  • Lecture 5 1up.
  • Lecture 5 4up.
  • Reading Mitchell pp. 27-33. Start looking at Mitchell chapter 4.
  • Reading Whitley's GA tutorial (notes chapter 2) up to the start of section 4.1.
• Lecture 6. Evolving Neural Networks
  • Lecture 6 1up.
  • Lecture 6 4up.
  • Reading Mitchell section 2.3 and Matt Quinn's paper on Evolving communication without dedicated communication channels, 6th European Conference on Artificial Life (ECAL 2001), LNAI 2159, J. Kelemen and P. Sosik (eds.), 357-366, Springer-Verlag, Heidelberg 2001. This whole proceedings is available online from within the ed domain and has lots of interesting papers in it. Details of the network encoding are in A comparison of approaches to the evolution of homogeneous multi-robot teams, in Proc. Congress Evolutionary Computation (CEC01), Seoul, 128-135, IEEE Press, 2001.
• Lecture 7. Building Block Hypothesis and Hitch-hiking
  • Lecture 7 1up.
  • Lecture 7 4up.
  • Reading Mitchell sections 4.1 and 4.2.
• Lecture 8. FlatLand: an example of an evolved controller.
  • Lecture 8 1up.
  • Lecture 8 4up.
  • Reading: Georgios Yannakakis et al.'s paper on Performance, robustness and effort cost comparison of machine learning mechanisms in FlatLand, Proc. 11th Mediterranean Conf. on Control and Automation MED'03, IEEE, June 2003.
• Lecture 9. Genetic Programming
  • Lecture 9 1up.
  • Lecture 9 4up.
  • Reading: GAGP lecture notes Chapter 9 and John Koza's 1990 paper Genetic Programming: a paradigm for genetically breeding populations of computer programs to solve problems, Stanford CS Technical Report STAN-TR-CS 1314, or his 1992 book. Skim the Koza material (it's a long report) for details of the material covered in the lecture and for extra examples. Also look at Koza's extensive GP pages .
• Lecture 10. Genetic Planning: another GA/GP example in use.
  • Lecture 10 1up.
  • Lecture 10 4up.
  • Reading: Henrik Westerberg's Genplan paper.
• Lecture 11. Designing a GA
  • Lecture 11 1up.
  • Lecture 11 4up.
  • Reading: Mitchell Chapter 5.
• Lecture 12. Designing a GA
  • Lecture 12 1up.
  • Lecture 12 4up.
  • Reading: Mitchell Chapter 5.
  • Reading: Section 10 of Whitley tutorial (notes Chapter 2).
• Lecture 13. Ant Colony Optimisation
  • Lecture 13 1up.
  • Lecture 13 4up.
  • Reading: Dorigo M., V. Maniezzo & A. Colorni (1996). Ant System: Optimization by a colony of cooperating agents. IEEE Transactions on Systems, Man, and Cybernetics-Part B, 26, 1, 29-41, via Marco Dorigo's web page.
  • And: Dorigo M. & L.M. Gambardella (1997). Ant Colony System: A Cooperative Learning Approach to the Traveling Salesman Problem. IEEE Transactions on Evolutionary Computation, 1, 1, 53-66.
  • Check out the TSPLIB in Heidelberg if you want to look at standard problems
• Lecture 14. Ant Colony Optimisation and the Bin Packing Problem
  • Lecture 14 1up.
  • Lecture 14 4up.
  • Reading: John Levine and Frederick Ducatelle (2004). Ant Colony Optimisation and Local Search for Bin Packing and Cutting Stock Problems. Journal of the Operational Research Society 55(7), 705-716 (2004). Also look at Frederick's home page to see his dissertation and other papers on Bin Packing, Cutting Stock and Mobile Ad Hoc Networks.
• Lecture 15. Using Genetic Programming to Evolve Policies in Planning.
  • Lecture 15 pdf.
  • Reading: John Levine and David Humphreys's paper on Learning Action Strategies for Planning Domains Using Genetic Programming, Proc. 4th European Workshop on Scheduling and Timetabling (EvoSTIM 2003), LNCS 2611, 688-699, Springer-Verlag, Heidelberg, 2003.
• Lecture 16. Interactions Between Learning and Evolution
  • Lecture 16 1up.
  • Lecture 16 4up.
  • Reading: David Ackley and Michael Littman's paper on Interactions Between Learning and Evolution, and Theodoros Damoulas's Evolving a Sense of Valency (for reference - no need to read it all).
  • Reading: Mitchell section 3.1.
• Lecture 17. Selection Revisited
  • Lecture 17 1up.
  • Lecture 17 4up.
  • Reading: Mitchell Chapter 4.