Editing Nonlinear dimensionality reduction (section)

=== Self-organizing map ===

The [[self-organizing map]] (SOM, also called ''Kohonen map'') and its probabilistic variant [[generative topographic mapping]] (GTM) use a point representation in the embedded space to form a [[latent variable model]] based on a non-linear mapping from the embedded space to the high-dimensional space.<ref>{{cite book |last=Yin |first=Hujun |chapter-url=http://pca.narod.ru/contentsgkwz.htm |chapter=3. Learning Nonlinear Principal Manifolds by Self-Organising Maps |editor1-first=A.N. |editor1-last=Gorban |editor2-first=B. |editor2-last=Kégl |editor3-first=D.C. |editor3-last=Wunsch |editor4-first=A. |editor4-last=Zinovyev |title=Principal Manifolds for Data Visualization and Dimension Reduction |series=Lecture Notes in Computer Science and Engineering |volume=58 |publisher=Springer |date=2007 |pages=68–95 |isbn=978-3-540-73749-0}}</ref> These techniques are related to work on [[density networks]], which also are based around the same probabilistic model.