Joint Estimation of Trajectory and Dynamics from Paired Comparisons (bibtex)
by , , and
Abstract:
Recent literature has developed methods for localizing a low-dimensional vectorwfrom paired comparisons of the form ``w is closer to p than q,'' where p and q are selected froma fixed set of landmark points and w does not change overtime. In this work, we consider a time-varying extension of this problem, in which w evolves according to some unknown dynamics model. We consider the task of actively selecting informative paired comparisons between landmark points to jointly estimate the state trajectory and identify the true dynamics model from a finite set of candidate models. Leveraging information-theoretic insights, we propose selecting pairs that simultaneously maximize information gain about both the trajectory and dynamics model, and propose a Bayesian method for tracking and system identification. We demonstrate the efficacy of our approach with numerical simulations, showing that our method is able to jointly estimate state dynamics and identify the correct dynamical model.
Reference:
Joint Estimation of Trajectory and Dynamics from Paired ComparisonsG. Canal, M. O’Shaughnessy, C. Rozell and M. Davenport. In International Workshop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP), December 2019. Submitted
Bibtex Entry:
@InProceedings{canal.19d,
      author = 	 {Canal, G. and  O’Shaughnessy, M. and Rozell, C. and Davenport, M.},
      title = 	 {Joint Estimation of Trajectory and Dynamics from Paired Comparisons},
     booktitle =	 {International Workshop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP)},
     year =	 2019,
	 address = {Guadeloupe, West Indies},
	 month = dec,
	 note = {Submitted},
	 abstract = {Recent literature has developed methods for localizing a low-dimensional vectorwfrom paired comparisons of the form ``w is closer to p than q,'' where p and q are selected froma fixed set of landmark points and w does not change overtime.  In this work, we consider a time-varying extension of this problem, in which w evolves according to some unknown dynamics model.  We consider the task of actively selecting informative paired comparisons between landmark points to jointly estimate the state trajectory and identify the true dynamics model from a finite set of candidate models. Leveraging information-theoretic insights, we propose selecting pairs that  simultaneously  maximize  information  gain  about  both the trajectory and dynamics model, and propose a Bayesian method for tracking and system identification.   We demonstrate  the  efficacy  of  our  approach  with  numerical  simulations, showing that our method is able to jointly estimate state dynamics and identify the correct dynamical model.}
  }
Powered by bibtexbrowser