3 ms·
word2vec is an algorithm to produce meaningful "word embeddings", which is a vector representation in a usually high-dimensional space. t-SNE is a dimensionalit
by cjauvin 11y ago
word2vec is an algorithm to produce meaningful "word embeddings", which is a vector representation in a usually high-dimensional space. t-SNE is a dimensionality-reduction algorithm. Both can be used together, as they serve different purposes.
- jknz 11y agoOne could argue that Word embeddings are also dimensionality-reduction techniques: Words live in an infinite dimensional space, and the embeddings is a finite-dimensional projection of this infinte dimensional space.
- mrdrozdov 11y agoI think of word vectors in the opposite light. Words stored in a dictionary have 1-dimension (their index), making comparisons more or less random. Word vectors augment the information you have about a word by continually examining the context that the word appears in a corpus of text.
- josh11b 11y agoIn fact, the typical way to visualize word2vec embeddings is t-SNE.