Dr Theo Michelot

Dr Theo Michelot

Honorary Lecturer

Researcher profile

208 OB


Research areas

I work on developing statistical methods and software, usually for the analysis of ecological data. Most of my work has been motivated by the modelling of animal movement, to better understand animals' behaviour and space use. For this purpose, I have worked various time series models, including state-space models, hidden Markov models, and stochastic differential equations.

Lately, I have been particularly interested in two questions:

  • Can we capture the time-varying behaviour of animals using models with parameters that are formulated as flexible (nonparametric) functions of time-varying covariates? This is for example crucial to identify the environmental drivers of animal behaviour, or to understand the impact of human activity on animals' activity.
  • Can we specify models of animals' short-term movement decisions that correctly predict their long-term spatial distributions? This is a long-standing theoretical problem in ecology, and stationary stochastic processes offer an interesting option to link the two temporal scales.

Selected publications


See more publications