bkinghor

main topics

• Overview of evolutionary algorithms
      Problems not for evolutionary algorithms
      Problems for evolutionary algorithms
  Architecture of an evolutionary algorithm in a nutshell
  Deterministic versus stochastic methods
  Types of optimization strategies – how to solve a problem
  
• Non EC stochastic heuristic methods
  Types of Evolutionary Algorithms and problem matching
      Evolutionary Programming (EP)
      Evolution Strategies (ES)
      Genetic Algorithms (GAs)
      Genetic Programming (GP)
      Learning Classifier Systems (LCS)
      Differential Evolution (DE)
      Gene Expression Programming
  
• Computational representation and implementation
      Population
      Selection
      Generation structure
      Fitness
      Search operators
      Crossover
      Mutation
  
• Fitness and Objective Functions
  Bloat
  Managing constraints
      How to apply constraints
      Hard and soft constraints
  Changing goal posts
      How to manage change of direction
      Opportunities for changing direction
  Multicriteria optimization
  Parallelization
  Self-evolving parameters
  Hybrid Evolutionary Algorithms
  Hybrid Algorithms
  Diagnosing convergence
      Criteria for stopping

why use EC?

This course aims to share some knowledge and skills that have greatly benefitted the presenters. The development of evolutionary algorithms is not a hard science - there are tips and tricks, and opportunities for tweaking.  There is an art to developing and using these algorithms, and much of the content of this course relates to personal experience and discoveries made in applications related to quantitative genetics and bioinformatics.  However applicability is very wide indeed. The process of evolution has brought about great complexity in a simple manner.  Biology is mind-bogglingly complex, even to our shallow level of understanding.  How could all of this have happened?  How could massive sequence and other components be derived to result in life as we see it? Working with evolutionary algorithms to solve complex problems gives some intriguing insight - it helps to give realization that indeed complex outcomes can come from a relatively simple process. As practitioners setting out to solve complex problems, we are not constrained to the systems of replication and propagation that we see in biology.  We can engineer our own systems to give faster and more robust evolutionary change towards solutions for the prevailing problem. We can target our outcomes in a more controlled yet flexible manner than in biological evolution.  We can explore the changes that could be made, and use this information to help set target outcomes for our prevailing problems - we can make a steering wheel.

To get started just download practicals, slides and course notes from the DOWNLOAD link above.

some examples of biological applications using evolutionary computation

Gondro, C., E. M. Strucken, H. K. Lee, K. D. Song and S. W. Lee (2014). Selection of SNP Panels for Parentage Testing. Proceedings of the 10th World Congress on Genetics Applied to Livestock Production.

Gondro, C. and P. Kwan (2012). "Parallel evolutionary computation in R". Multidisciplinary Computational Intelligence Techniques: Applications in Business, Engineering and Medicine. S. Ali, N. Abbadeni and M. Batouche. IGI Global: 351-377.

Gondro, C. and B. P. Kinghorn (2008). Optimization of cDNA microarray experimental designs using an Evolutionary Algorithm. IEEE Transactions on Computational Biology and Bioinformatics 5(4): 630-638.

Gondro, C. and B. P. Kinghorn (2007). Solving complex problems with evolutionary computation. Proceedings of the 17^th Conference of the Australian Association for the Advancement of Animal Breeding and Genetics (invited). 17: 272-279.

Gondro, C. and B. P. Kinghorn (2007). A simple genetic algorithm for multiple sequence alignment. Genetics and Molecular Research 6(4): 964-982

acknowledgements

Thanks to all the students who took the course and provided valuable input into its structure and contents.

assistance

If you have any comments, questions, suggestions or find any bugs please let me know. To organize a short couse at your institution just contact me at the email address below.