What are the best practices when using Hidden Markov Models for movement data in terms of daily GPS fix frequency? Are there any? I've seen data being used in the Vignettes of moveHMM R package, for example, which use many daily, and few daily GPS fixes.
Although I know more would be better to draw accurate inferences, does anyone have an idea of what the minimum would be for daily GPS fixes in which HMM could still accurately account for behavioral state changes?
Article Hidden Markov Models: The Best Models for Forager Movements?
http://papers.nips.cc/paper/669-hidden-markov-model-induction-by-bayesian-model-merging.pdf
Article A New Method for Inferring Hidden Markov Models from Noisy T...
Article Effective Identification of Conserved Pathways in Biological...
Article Profile Hidden Markov Models for the Detection of Viruses wi...
https://pdfs.semanticscholar.org/ac73/7af2498052efc1736a3624646f9c118be1b5.pdf
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.957.202&rep=rep1&type=pdf
Conference Paper Learning Clinical Pathway Patterns by Hidden Markov Model.
Some more about Markov Chains usability: https://towardsdatascience.com/introduction-to-hidden-markov-models-cd2c93e6b781Chapter Wireless Channel Model with Markov Chains Using MATLAB
- Badges
- Science topic
- Similar topics
- Names