Animal trajectory imputation and uncertainty quantification via deep learning

Environmetrics
By: , and 

Links

Abstract

Imputing missing data in animal trajectories is crucial for understanding animal movements during unobserved periods. However, the traditional methods, such as linear interpolation and the continuous-time correlated random walk model, are often inadequate to capture the complexity of animal movements. Here, we develop a deep learning approach to animal trajectory imputation by a conditional diffusion model. Unlike the traditional methods, our deep learning method uses observed data and external covariates to impute missing positions along an animal trajectory, capturing periodic patterns and the influence of covariates, which leads to more accurate imputations. In a case study of imputing deer trajectories, our method not only provides more accurate deterministic imputations than existing approaches but also achieves uncertainty quantification through probabilistic imputation.

Suggested Citation

Yao, K., McGahan, I.P., Zhu, J., Storm, D.J., Walsh, D.P., 2025, Animal trajectory imputation and uncertainty quantification via deep learning: Environmetrics, v. 36, no. 6, e70027, 15 p., https://doi.org/10.1002/env.70027.

Publication type Article
Publication Subtype Journal Article
Title Animal trajectory imputation and uncertainty quantification via deep learning
Series title Environmetrics
DOI 10.1002/env.70027
Volume 36
Issue 6
Publication Date July 23, 2025
Year Published 2025
Language English
Publisher Wiley
Contributing office(s) Coop Res Unit Seattle
Description e70027, 15 p.
Additional publication details