In ML the term "embedding" gets tossed around a lot and the term basically means the construction of a function that takes a high-dimensional vector to a low-dimensional vector in such a way that the high-dimensional vector can be recovered by the low-dimensional vector. (...at least in the case of autoencoders, not sure if word embedding cares about reconstructing the word.)
Embedding was also a word that I encountered in topology and differential geometry. https://en.wikipedia.org/wiki/Embedding
Does there exist any correspondence between the usage of these terminologies between math and ML?