skip to content
 

Spatiotemporal modelling and parameter estimation of anisotropic particle trajectories

Presented by: 
Adam Sykulski Lancaster University, University College London
Date: 
Tuesday 27th March 2018 - 11:00 to 12:00
Venue: 
INI Seminar Room 2
Abstract: 
Trajectories of moving objects are collected everywhere and in massive volumes. In the first part of this talk I will present a framework for stochastically modelling such trajectories in time over two-dimensional space. I will show an application to modelling fluid particles in ocean turbulence, where trajectories are typically anisotropic due to spherical dynamics. In the second part, I will discuss computationally-efficient parameter estimation for massive datasets, where we propose an important modification to the Whittle likelihood to significantly reduce bias at no additional computational cost. We extend these estimation methods to spatiotemporal trajectories, such that we can estimate parameters that reveal the nature of the anisotropic structure.



The video for this talk should appear here if JavaScript is enabled.
If it doesn't, something may have gone wrong with our embedded player.
We'll get it fixed as soon as possible.
University of Cambridge Research Councils UK
    Clay Mathematics Institute London Mathematical Society NM Rothschild and Sons