IS = { zkontrolovano 07 Dec 2003 },
  UPDATE  = { 2003-12-04 },
author =      {De Ridder, Dick and Franc, Vojt{\v e}ch},
title =       {Robust Manifold Learning},
institution = {Center for Machine Perception, K13133 FEE
               Czech Technical University},
address =     {Prague, Czech Republic},
year =        {2003},
month =       {April},
type =        {Research Report},
number =      {{CTU--CMP--2003--08}},
issn =        {1213-2365},
pages =       {36},
figures =     {10},
authorship =  {90-10},
psurl       = {[deRidder-TR-2003-08.pdf]},
project =     {MIRACLE ICA1-CT-2000-70002, MSM 212300013},
annote =    {Probabilistic subspace mixture models, as proposed over the
               last few years, are interesting methods for learning image
               manifolds, i.e. nonlinear subspaces of spaces in which
               images are represented as vectors by their grey-values.
               Their lack of a global mapping can be remedied by a
               recently developed method based on locally linear
               embedding, called locally linear coordination. However, for
               many practical applications, where outliers are common,
               this method lacks the necessary robustness. Here the idea
               of robust mixture modelling by $t$-distributions is
               combined with probabilistic subspace mixture models. The
               resulting robust subspace mixture model is shown
               experimentally to give advantages in density estimation and
               classification of image data sets. It also solves the
               robustness problems of locally linear coordination, by
               introducing a weighted reformulation of the embedding step. },
keywords =    {manifolds, mixture models, coordination, locally linear embedding},