Developing novel hidden Markov models from a data science perspective to inform seabird conservation

了解动物运动和确定影响动物运动的行为对有效的保护管理至关重要。行为通常是从运动数据中推断出来的，很少得到验证，因为很难直接观察和记录在海洋环境等偏远和恶劣地形中影响动物运动的行为。使用联合自然保护委员会（JNCC）提供的数据，使用隐马尔可夫模型（hmm）从运动数据推断出的行为的有效性进行了检验。此外，动物运动数据的时间分辨率可以影响对动物行为的推断。例如，以较低的时间分辨率获得的动物运动数据可能无法捕捉到行为，只有在更精细的时间尺度上才能获得，这需要花费采样时间和GPS电池寿命。我们从Nyquist-Shannon采样定理（NSST）的角度来研究这个问题，该定理用于选择动物运动轨迹中的时间分辨率，这样就不会丢失有关潜在行为过程的信息。

Understanding animal movement and identifying the behaviours that influence animal movement is crucial to inform effective conservation management. Behaviours are often inferred from movement data and rarely validated because it is difficult to directly observe and record behaviours influencing animal movement in remote and hostile terrain such as the marine environment. Using data provided by the Joint Nature Conservation Committee (JNCC), the validity of behaviours inferred from movement data using hidden Markov models ( HMMs) is examined. Also, the temporal resolution of animal movement data can impact inference on animal behaviour. For instance, animal movement data obtained at a lower temporal resolution may not capture behaviours only obtainable at a finer temporal scale that comes at a cost sampling duration and GPS battery life. We investigate this problem from the optic of the Nyquist-Shannon sampling theorem (NSST) useful for selecting temporal resolutions in the animal movement trajectories such that there is no loss of information about the underlying latent behavioural process.