Manifold learning for multi-dimensional auto-regressive dynamical   models






Fabio Cuzzolin
Machine Learning for Vision-based Motion Analysis, 2010
Abstract

We present a general differential-geometric framework for learning distance functions for dynamical models. Given a training set of models, the optimal metric is selected among a family of pullback metrics induced by the Fisher information tensor through a parameterized automorphism. The problem of classifying motions, encoded as dynamical models of a certain class, can then be posed on the learnt manifold. In particular, we consider the class of multidimensional autoregressive models of order 2. Experimental results concerning identity recognition are shown that prove how such optimal pullback Fisher metrics greatly improve classification performances.
sp 
Download
PDF  Zipped Postscript

BibTeX entry

@incollection{cuzzolin10springer, 
  AUTHOR = "Fabio Cuzzolin", 
  TITLE = "Manifold learning for multi-dimensional auto-regressive dynamical models",
  BOOKTITLE = "Machine Learning for Vision-based Motion Analysis", 
  EDITOR = "L. Wang and G. Zhao and L. Cheng and M. Pietikäine", 
  PUBLISHER = "Springer", 
  YEAR = "2010" 
}