On the Use of Human-Guided Evolutionary Algorithms for Tackling 2D Packing Problems - Foundations on Natural and Artificial Computation - page 359

Fig. 4. Best solution found in the 49-object instance with the automated EA (left),
and the human-guided EA (right)
5 Conclusions
This work has presented an initial progress report on the use of human-guided
evolutionary algorithms for solving 2D packing problems. In this sense, the re-
sults are encouraging and hint the potential for user-centric techniques in this
domain. A point of caution when interpreting the results is required though.
Besides using a more ample test suite to analyze the scalability of this kind of
techniques, it is required to analyze in a deeper way the changes the user in-
teraction is exerting on the search dynamics and population diversity. It may
be the case that these effects can be reverse-engineered and integrated into an
automated EA approach. Then again, this would indicate the usefulness of the
human guide, at least from a designer's point of view. Future work will try to
confirm this.
Acknowledgements. This work is supported by project NEMESIS (TIN-2008-
05941) of the Spanish Ministerio de Ciencia e Innovacion, and project TIC-6083
of Junta de Andaluc ıa.
References
1. Dyckhoff, H.: A typology of cutting and packing problems. European Journal of
Operational Research 44, 145-159 (1990)