Original link:tecdat.cn/?p=17592
Recently, we developed a solution using a hidden Markov model and were asked to explain the solution.
HMM is used to model data sequences, whether derived from continuous or discrete probability distributions. They are relevant to state space and Gaussian mixture models because they aim to estimate the states that give rise to observations. The state is unknown or “hidden”, and the HMM attempts to estimate the state, similar to an unsupervised clustering process.
Video: An example of the Hidden Markov HMM model in R