Machine Learning Idea – 3D Spherical Boundary Map As Decision Basis?

I was wondering.

Why settle with 2D maps of boundaries. Philosophically it should be possible to have 3D spherical maps. Like the earth globe with clouds as boundaries.

Maybe not all objects are suitable as basis for boundaries. But I think 2D rectangles and spheres should the first ones to try.

What possibilities does this have?

(Visited 4 times, 1 visits today)