Machine Learning in Architectural Design

Course Description: An overview of machine learning tools and techniques that can be used in architectural design process, applying machine learning workflows as generative experiments, selecting and using some of the Artificial Intelligence tools depending on the needs of architectural design, collecting and preparing the dataset in groups, selecting and training some machine learning model and finally generating designs using the created model. Course Objectives: Acquire an overview of machine learning models and learn the potentials and weaknesses of each model. Learn how deep neural networks work and how to use them in architectural context. Obtain the skill of selecting the suitable machine learning platform and the skill of adapting the platform for the design purpose. Practice on obtaining a dataset, cleaning and structuring it and converting it to a machine interpretable entities to be used in generative process. Create architectural designs by using machine learning approaches.

weekly

29/01/2024

Master students who successfully pass this course gain knowledge, skills and proficiency in
the following subjects:
1. General knowledge of machine learning models
2. Basics of Neural networks and their applications in architecture
3. Design of examples that uses machine learning tools for design generation
4. Predict use of spaces by using machine learning
5. Create a report for the final project that clearly communicates the process and
methodology used and future possibilities.
6. Present the report to a crowd that describes the whole process from concept to the result
with the emphasis of the potentials of the machine learning workflows.

None

42 h - 14 weeks

7.5

To be specified

5

İstanbul Teknik Üniversitesi

Graduate School

Architectural Design Computing