| |||||||||
|
|
| Architectural Design with Machine Learning, 2024 |
| Wahlfach HS |
|
Deutsche Übersetzung folgt demnächst... This elective course presents machine-learning methods for data-driven design exploration in architectural design, and how to leverage them in combination with the parametric modelling paradigm. The students learn how to harness parametric models for data exploration and how to augment the design process with project-specific generative deep learning models. They acquire basic understanding of the underlying methods and can implement them in their design tasks in architecture, urban planning, engineering etc. The course will cover the following topics: data exploration (analytics and visualisations), forward and inverse design problems (enhancing parametric modelling with machine learning), basics of machine learning (feed-forward networks, backpropagation), and generative models with special focus on autoencoders. The course consists of lectures providing a theoretical background followed by hands-on practical sessions with coding exercises in Python and Grasshopper. In parallel, the students will work in small groups on a semester project, in which they apply the presented data-exploration and inverse design methods to a design task of their choice. Building upon the provided framework (in Grasshopper and Python), students will generate custom datasets and train project-specific models, and then use them for concept-phase design exploration.
|
| Credits: |
|
Gramazio Kohler Research, ETH Zürich Studenten: |
| Copyright 2024, Gramazio Kohler Research, ETH Zurich, Switzerland | ||
| ||
|
Gramazio Kohler Research
Professur für Architektur und Digitale Fabrikation ETH Zürich HIB E 43 Stefano-Franscini Platz 1 / CH-8093 Zürich +41 44 633 49 06 | Follow us on: Vimeo | Instagram | |