Semi Markov

Semi-Markov models extend traditional Markov models by allowing state durations to follow arbitrary distributions, enabling more realistic modeling of time-dependent processes. Current research focuses on improving the robustness and interpretability of these models, particularly within Hidden Semi-Markov Models (HSMMs), through techniques like robust hierarchical Dirichlet processes and model-based tree structures. These advancements find applications in diverse fields, including natural language processing, transportation research (e.g., analyzing driving patterns), healthcare (e.g., predicting mortality risk), and building occupancy modeling, offering improved accuracy and explainability in analyzing sequential data.