Learning Embeddings into Entropic Wasserstein Spaces

    Charlie Frogner
    Charlie Frogner
    Farzaneh Mirzazadeh
    Farzaneh Mirzazadeh

    ICLR, 2019.

    Cited by: 0|Bibtex|Views26|Links
    EI

    Abstract:

    Euclidean embeddings of data are fundamentally limited in their ability to capture latent semantic structures, which need not conform to Euclidean spatial assumptions. Here we consider an alternative, which embeds data as discrete probability distributions in a Wasserstein space, endowed with an optimal transport metric. Wasserstein spa...More

    Code:

    Data:

    Your rating :
    0

     

    Tags
    Comments