Lesson 25b Introduction to State Space Models
Boston University EE509 "Applied Environmental Statistics" Course: In this lecture, the second in our discussion of time-series analysis, we introduce the State Space modeling framework (a.k.a. Hidden Markov models), a powerful framework that allows us to fit dynamic models, and partition observation and process errors, by treating the true, unobserved state as a series of latent variables. We illustrate this framework starting with a simple Random Walk state-space model. http://people.bu.edu/dietze/Bayes2020...