What are Interactive Genetic Algorithms (IGAs)

As discussed earlier, evolutionary algorithms draw analogies from natural evolutionary processes, effectively solving diverse problems. Since the 1980s, GAs have been used in optimisation and classification tasks, proving adept at finding optimised solutions using definite fitness values. However, GAs struggle to incorporate subjective goals crucial in creative fields such as art, design, music, and architecture. In these areas, assigning a fitness value is difficult as evaluations are heavily reliant on human subjective judgement. Furthermore, in creative fields, integrating the designer's input into the evolutionary process is essential for guiding the design based on subjective judgement.

IGAs involve the architect directly in the design process, allowing them to shape the evolution of solutions based on personal preferences. This approach is key to creating effective, user-centered evolutionary systems, as it derives fitness values directly from the designer, influencing the choices for future generations. Additionally, IGAs blend subjective elements and human interaction, offering architects an interactive and iterative exploration beyond mere optimisation. Researchers have defined the principle workflow of IGA as a synthetic approach that embeds a human in the evolutionary loop and allows the algorithm to select the targeted individuals according to user's evaluation or selection.