Skip to main content
Fig. 1 | Movement Ecology

Fig. 1

From: A hierarchical machine learning framework for the analysis of large scale animal movement data

Fig. 1

Detection of periodic activity pattern. (a) Coordinates of the simulated individual. The amplitude of the random movement increases for the same period each day. (b) Samples from the posterior. Using Markov chain Monte Carlo (MCMC) sampling we can sample from the posterior distribution over functions, hence we can infer the latent activity levels as a function of time of day. (c) Once we have taken sufficient samples we can calculate the posterior mean (red line) and 95% credible interval (shaded region). The true latent function is shown as a black dashed line

Back to article page