Emergent Syntax: Machine Learning for the Curation of Design Solution Space

The expanding role of computational models in the process of design is producing exponential growth in parameter spaces. As designers, we must create and implement new methods for searching these parameter spaces, considering not only quantitative optimization metrics but also qualitative features. This paper proposes a methodology that leverages the pa ern modeling properties of artificial neural networks to capture designers’ inexplicit selection criteria and create user-selection-based fitness functions for a genetic solver. Through emulation of learned selection patterns, fitness functions based on trained networks provide a method for qualitative evaluation of designs in the context of a given population. The application of genetic solvers for the genera on of new populations based on the trained network selections creates emergent high-density clusters in the parameter space, allowing for the identification of solutions that satisfy the designer’s inexplicit criteria. The results of an initial user study show that even with small numbers of training objects, a search tool with this configuration can begin to emulate the design criteria of the user who trained it.