Learn with Embeddings

word2vec

CBOW model: predicts target words from source context words.

Skip-gram model: predicts source context words from the target words.

FastText

t-Distributed Stochastic Neighbor Embedding (t-SNE)

PCA is a linear algorithm. It will not be able to interpret complex polynomial relationship between features. On the other hand, t-SNE is based on probability distributions with random walk on neighborhood graphs to find the structure within the data.

Local approaches seek to map nearby points on the manifold to nearby points in the low-dimensional representation. Global approaches on the other hand attempt to preserve geometry at all scales, i.e mapping nearby points to nearby points and far away points to far away points