t-Distributed Stochastic Neighbor Embedding

t-distributed stochastic neighbor embedding, or t-SNE, is a statistical method for visualizing high-dimensional data by giving each data point a location in a two or three dimensional map. It is a nonlinear dimensionality reduction technique built on the earlier stochastic neighbor embedding method developed by Geoffrey Hinton and Sam Roweis, with Laurens van der Maaten and Hinton proposing the t-distributed variant. It models each high-dimensional object as a point in low-dimensional space such that similar objects are placed near each other and dissimilar objects are placed far apart with high probability.