Fig. 4From: Hidden Markov Model based stride segmentation on unsupervised free-living gait data in Parkinson’s disease patientsHMM structure of the stride sub-model: A left-right Markov chain from state \(s_{0}\) to \(s_{n}\), with transition probability \(p_{n,n+1}\), and self transition probability \(p_{n,n}\). The sequence is enforced to always start with state \(s_{0}\) and end with \(s_{n}\)Back to article page