Introducing Machine Learning within an Interactive Evolutionary Design Environment

The overall research focuses on the development of an Interactive Evolutionary Design System (IEDS) [Parmee 2002] which integrates machine-based evaluation of engineering criteria and rulebased aesthetic criteria with the designer’s subjective aesthetic evaluation of design solutions. The research described in the paper relates to user-centric intelligent design systems and creativity in design. While research relating to artificial design environments is evident in the literature [Gero 2002], [Bentley 1999] there is little evidence of the integration of user evaluation, evolutionary search and exploration and machine learning. Furthermore, integration of aesthetic criteria within computerbased design has been limited to the development of theoretical models with little evidence of application based research [Moore et al. 1996; Saunders 2001]. The IEDS in this case addresses the conceptual design of bridges. For a more detailed description of the system as illustrated in Figure 1, the reader is directed to Machwe et. al. [2005]. The research decribed in the following sections concentrates upon the machine learning sub-system .