Title :
A Fast-Converging Algorithm for Nonlinear Mapping of High-Dimensional Data to a Plane
Author :
Niemann, Heinrich ; Weiss, Jürgen
Author_Institution :
Friedrich-Alexander-Universitat, Institut fur Mathematische Maschinen und Datenverarbeitung
Abstract :
An iterative algorithm for nonlinear mapping of high-dimensional data is developed. The step size of the descent algorithm is chosen to assure convergence. Steepest descent and Coordinate descent are treated. The algorithm is applied to artificial and real data to demonstrate its excellent convergence properties.
Keywords :
Cluster analysis; coordinate descent; dimensionality reduction; iterative algorithm; nonlinear mapping; steepest descent; unsupervised learning; Algorithm design and analysis; Convergence; Iterative algorithms; Iterative methods; Nonlinear distortion; Pattern analysis; Unsupervised learning; Cluster analysis; coordinate descent; dimensionality reduction; iterative algorithm; nonlinear mapping; steepest descent; unsupervised learning;
Journal_Title :
Computers, IEEE Transactions on
DOI :
10.1109/TC.1979.1675303
