The effect of user interaction mechanisms in multi-objective IGA

ABSTRACT

In this paper four mechanisms, fine and coarse grained fitness rating, linguistic evaluation and active user intervention are compared for use in the multi-objective IGA. The interaction mechanisms are tested on the ergonomic chair design problem. The active user intervention mechanism provided the best fitness convergence but resulted in the least diverse results. The fine grained evaluation provided a good blend of fitness convergence and diversity while the popular coarse grained discrete rating provided poor results. Linguistic evaluation resulted in poor qualitative fitness despite its fast speed of evaluation. The significant differences between interaction mechanisms show the need for further research.

References

1. Takagi, H. Interactive Evolutionary Computation: Fusion of the capabilities of EC Computation and Human Evaluation, Proceedings of the IEEE, 89:9, pp. 1275--1296, 2001.
2. Brintrup, A. M., Tiwari A., and Gao J., A Framework for Handling Qualitative and Quantitative Data in an Evolutionary Multiple Objective Design Space, Int. J. of Computational Intelligence, 1:4, pp. 282--289, 2004.
3. Brintrup, A. M., Handling Qualitativeness in Design Optimisation using Interactive and Multi-objective Genetic Algorithms, Ph.D. Thesis, Cranfield University, UK, 2007.
4. Kim, H. S. and Cho, S. B., Application of Interactive Genetic Algorithm to Fashion Design, Engineering Applications of Artificial Intelligence, 13:6, pp. 635--644, 2000.
5. Ohsaki, M., Takagi, H., and Ohya K., An Input Method Using Discrete Fitness Values for Interactive GA, J. of Intelligent and Fuzzy Systems, 6:1, pp. 131--145, 1998.
6. Takagi, H. and Kishi, K., On-line Knowledge Embedding for Interactive EC-based Montage System, Int. Conf. on Knowledge-Based Intelligent Information Engineering Systems (KES'99): Adelaide, Australia, pp. 280--283, 1999.
7. Unemi, T. SBART 2.4: an IEC Tool for Creating 2D Images, Movies, and Collage, Leonardo, 35:2, pp. 189--191, MIT Press, 2002.
8. Kamalian, R., Takagi, H., and Agogino, A., Optimized Design of MEMS by Evolutionary Multi-Objective Optimization with Interactive Evolutionary Computation, Genetic and Evolutionary Computation Conference (GECCO2004): Seattle, pp. 1030--1041, 2004.
9. Machwe, A., Parmee, I. C., and Miles, J., Integrating Aesthetic Criteria with a User-centric Evolutionary System via a Component-based Design Representation, Int. Conf. on Engineering Design (ICED2005), Melbourne, 2005.
10. Deb, K., Agrawal, S., Pratap, A., and Meyarivan, T., A Fast Elitist Non dominated Sorting Genetic Algorithm for Multi Objective Optimization: NSGA-2, Parallel Problem Solving from Nature (PPSN2000): Paris, pp. 858--862, 2000.
11. Brintrup, A., Ramsden, J., and Tiwari, A., An Interactive Genetic Algorithm Based Framework for Handling Qualitative Criteria in Design Optimization, J. of Computers in Industry, 58:3, pp.279--291, 2007.
12. Grosan, C., Oltean, M., and Dumitrescu, D. Performance metrics for multiobjective optimization evolutionary algorithms, Conference on Applied and Industrial Mathematics (CAIM2003): Oradea, 2003.
13. Tilley A. R., The Measure of Man and Woman: Human Factors in Design, Revised Edition, Wiley & Sons, 2002.

Index Terms

• Computing methodologies

  • Artificial intelligence

    • Search methodologies

      • Heuristic function construction