／var／logmarcus chiu

Non-Linear／Nonlinear Dimensionality Reduction (NLDR) - Manifold Learning

Created on · Last Modified on

Non-Linear/Nonlinear Dimensionality Reduction (NLDR) - Manifold Learning
  • refers to various related techniques that aim to project high-dimensional data onto lower-dimensional latent manifolds (using the assumption of the manifold hypothesis)
  • is a technique that finds a non-linear manifold within a higher-dimensional space
  • the goal of manifold learning is to project high-dimensional data onto lower-dimensional latent manifolds
  • are generalizations of linear decomposition methods used in dimensionality reduction (e.g. singular value decomposition and principal component analysis)

Subpages

Resources