Machine Learning for Design

Instructor: Panagiotis Michalatos, Nicholas Cassab

Group work/4

Tools: Python, Grasshopper

Time: 23Fall

Select two categories with distinct and different features, then input them as dataset into machine learning classifier, and then use ML's judgment for design selection, including siting, windowing, and structuring.

We chose classic American food and Japanese food, and you can see that there is a clear difference between the two in terms of color and presentation style.

We chose a location where there is a clear difference in the styles of the two sides, i.e., the site where the long warehouse is located, with the sea side representing the Japanese style and the land side representing the American food style.

We analyzed the sight lines at the corresponding locations, and roughly determined the style of window openings, from Japanese style to American style.

Structural styles were then tested and categorized, and it was found that standard piller and cave supports represented two structural styles at the extremes of the food spectrum.

In the end, the seaward side is supported by off-columns, the continental side is supported by its own structure, and the center is a transition.