Computational Explorations

Optimization, machine learning, and deep learning lie at the root of many of the most important scientific breakthroughs in the last decade. But should we expect artificial intelligence based on neural networks to soon replace human designers? Computational Explorations invites students to familiarize themselves with these advanced computational methods in the context of architectural design. Students will learn how to automate the search for good design candidates, analyze the resulting data, and make predictions from that data. Beyond practical skills in data science, familiarity with these methods will allow students to better understand and reflect on their impact on the architectural profession. The module focuses on performance-informed architectural design with building simulations, but students will be free to explore other applications of these methods as well. Assessment is based on several assignments and a final review (in small groups). The module assumes familiarity with Rhino/Grasshopper and Python programming (as taught in Computational Design).