The dimensionality of data isn't the dimensionality of the encoding medium (e.g. images -> 2D data). All data can be re-encoded with a different number of dimensions.
What matters is the dimensionality of the latent manifold. The structure of correlations in your data.
What matters is the dimensionality of the latent manifold. The structure of correlations in your data.
So it is definitely not the case that painting is 2D, music is 1D, and mathematics is infD. Mathematics, like computing, is typically encoded as manipulations on strings of symbols, so by that logic it would be a 1D problem.
Arguably all of these domains have the same "dimension", which is the intrinsic dimensionality of the human mind.
(I'm expecting pedantic, braindead replies about how mathematics is not a product of the human mind 😉)
