--- title: Hidden Semi-Markov Models (HSMMs) url: https://www.emergentmind.com/topics/hidden-semi-markov-models-hsmms type: topic --- # Hidden Semi-Markov Models (HSMMs) A hidden semi-Markov model (HSMM) is a probabilistic model for sequential data in which the latent process is a semi-Markov chain: the process remains in each hidden state for a random—explicitly modeled—duration before transitioning, emitting possibly complex, state-dependent observations. This explicit duration modeling lifts the geometric-duration constraint of classical hidden Markov models (HMMs), yielding substantially greater flexibility for modeling systems with history-dependent state persistence. Since their introduction, HSMMs have become foundational in time series segmentation, regime-switching modeling, survival analysis, structured sequence learning, and have seen technical developments in inference, representation, covariate conditioning, and nonparametric modeling. ## 1. Formal Model Structure and Duration Modeling In an HSMM, the latent process $S_t$ (discrete time) or $X(t)$ (continuous time) resides in some state $i\in\{1,\dots,K\}$ for a state-dependent random "dwell time" or sojourn, whose distribution is specified a priori or learned from data. At the end of the sojourn, the chain makes an instantaneous transition to a new state $j$ (with $j\ne i$), with transition probabilities independent of the sojourn duration unless generalized. The key distinction from an HMM is that the dwell-time distribution $d_i(\cdot)$ for each state $i$ is arbitrary—not constrained to be geometric (discrete time) or exponential (continuous time). This non-Markovianity introduces additional dependencies in the latent process, so that the transition probability depends not only on the current state but how long the chain has occupied it. The state-dwell distribution is often specified parametrically: e.g., shifted Poisson, negative-binomial, Weibull, or nonparametric (as in the empirical or HDP-HSMM setting). In covariate-extended (nonhomogeneous) HSMMs, the hazard or duration law may be linked to time-varying covariates, for instance via proportional hazards: \[ g\bigl(q_i(d|X_t)\bigr) = \eta_{0,i}(d) + \beta_i^{\top} X_t, \] where $g$ is typically the complementary log-log link, $\eta_{0,i}(d)$ is a baseline, and $\beta_i$ a regression vector [2312.14719]. The emission process is conditionally independent noise given the current regime, often allowing structured or high-dimensional likelihoods, such as VAR processes, toroidal densities, or mixtures. The complete-data likelihood for a sequence $(Y, S, D)$ factors into initial, transition/dwell, and emission components. ## 2. Inference Algorithms and Computational Representation Likelihood-based inference in HSMMs requires summation over all possible segmentations aligned with the observable data, with forward-backward procedures generalized to handle explicit duration modeling. The forward (and backward) variables must index both the latent state and the current sojourn's age or remaining duration. For discrete time, these recursions are quadratic in number of states $K$ and linear in support of the duration distribution $M$: \[ \alpha_t(j, d_j) = \sum_{i\ne j} \sum_{d_i} \alpha_{t-d_j}(i, d_i) \cdot a_{(i,d_i)\to(j,d_j)} \cdot b_{j, d_j}(o_{t-d_j+1:t}), \] with analogous backward recursions [1608.06954]. The EM algorithm is the canonical approach for frequentist learning: - **E-step:** compute expectations over the latent (augmented) state/duration indicators using the generalized forward-backward recursions. - **M-step:** maximize expected complete-data log-likelihood; transition probabilities, duration distribution parameters, and emission parameters typically remain in closed form or are solved by weighted regression and density estimation, e.g., via GLMs for proportional hazards models [2312.14719], or mixture models for emissions. For models with covariate-dependent duration laws, the M-step for hazard regression is a weighted binomial log-likelihood, enabling use of generalized linear models, possibly with complementary log-log link [2312.14719]. For the Bayesian paradigm, fully specified hierarchical models allow Gibbs or Metropolis-within-Gibbs posteriors, with latent state sequence and durations sampled via block updates or by a dynamic-programming recoding and Viterbi-like recursions [2010.10739, 2109.09949]. In high-frequency data scenarios exhibiting residual dependence, subsampling in the emission/observation parameter update step has been shown to effectively control estimator variance [2109.09949]. ## 3. Model Extensions: Nonhomogeneous, Interval, and Nonparametric HSMMs HSMMs have been generalized along several axes: - **Covariate-modulated (nonhomogeneous) HSMMs:** Duration distribution parameters (hazards or rates) can depend on covariates, e.g., via proportional hazard or log-linear models. Both parametric and nonparametric specifications exist, with efficient EM or Bayesian inference adapted for covariate-linked hazards [2312.14719, 2010.10739, 2109.09949, 2405.13553]. - **Interval modeling:** The interval-state HSMM (IS-HSMM) introduces explicit "interval" states for periods of inactivity or missing data, while the interval-length probability HSMM (ILP-HSMM) separately models the distribution of gap lengths, allowing richer representation of silence or inactivity [1608.06954]. - **Bayesian nonparametric HSMMs:** The HDP-HSMM allows an unbounded state space with a hierarchical Dirichlet process prior over transition measures, integrating explicit-duration modeling with nonparametric complexity control and MCMC algorithms for efficient posterior inference [1203.3485]. - **Higher-order HSMMs:** Extensions in which the latent sequence is a higher-order Markov process; permits multi-step dependencies in state transitions and complex, non-renewal patterns in hidden labels [2002.05272]. ## 4. Applications and Empirical Results HSMMs have been applied to diverse empirical problems: - **Physical and environmental time series:** Covariate-modulated duration modeling enables detailed segmentation of physiological or environmental signals, such as heart rate of athletes and phycocyanin concentration in lakes, with substantial improvements in interpretability and detection of external control factors [2010.10739, 2109.09949]. - **Movement and behavior analysis:** Explicit-duration HSMMs and their autoregressive variants improve segmentation accuracy in human gesture modeling, animal behavior classification, and wearable device streams, particularly when dwell time distributions are non-geometric or emissions are highly overlapping [2302.05347, 2105.11490]. - **Speech and sequence segmentation:** In synthetic and real datasets including Morse code clustering and speaker diarization, HDP-HSMMs have shown substantial gains in correct state recovery and segmentation by using explicit dwell-time modeling [1203.3485]. - **Regime-switching in financial time series:** HSMMs with AR emission structures and flexible dwell times outperform HMMs for volatility clustering and regime detection, with particle filter-based sequential Bayesian algorithms enabling tractable inference in long sequences [2301.10494]. - **Robotics and imitation learning:** HSMMs support segmentation of complex manipulation demonstrations, with task-parameterized and latent-space variants encoding geometric and subspace invariances central to few-shot robot learning [1811.07489]. ## 5. Computational Strategies and Algorithmic Considerations The primary computational bottleneck in HSMMs is the explicit duration indexing in forward-backward and Viterbi-type recursions, which introduces an $O(M)$ factor in both run time and memory scaling. Practical implementations use: - **State-space expansion:** Truncating the maximum dwell time $M$ sufficiently above the empirical maximum ensures accurate marginalization while controlling computational cost. Representing the semi-Markov chain as a Markov chain on an augmented (micro-state) space enables efficient algorithms, with block-matrix structure exploited for sparse or structured representation [2312.14719, 2006.09061, 2405.13553]. - **Spectral inference:** Moment-based spectral algorithms can recover likelihoods and estimate model parameters efficiently in the fully discrete case by leveraging joint counts and low-rank tensor structures, with computational complexity logarithmic in maximum state persistence [1407.3422]. - **Sequential Monte Carlo:** For Bayesian inference with long or streaming data, particle filter-based SMC and Particle Gibbs methods provide unbiased inference and are often orders of magnitude faster than full-recursion-based approaches, at a cost of Monte Carlo error [2301.10494]. In continuous time, semi-Markov processes generalize the discrete formulation, and recent work extends forward-backward and Viterbi methods to fully continuous settings via integro-differential equations and probability current representations [2210.09058]. ## 6. Theoretical Properties, Model Identifiability, and Information Complexity The statistical structure of HSMMs enables modeling of non-memoryless dwell times, but introduces identifiability challenges relating to the joint estimation of emission, transition, and duration distributions. Bayesian regularization, priors, and non-local constraints (such as those enforcing separation between geometric and non-geometric dwell families) are central to robust inference and model selection [2302.05347, 2006.09061]. The information-theoretic properties of HSMMs have been formalized in the context of causal-state representations: the minimal maximally predictive process representations ("ε-machines") for unifilar HSMMs yield closed-form expressions for entropy rate, statistical complexity, excess entropy, and differential information anatomy rates, elucidating the memory and predictability structure of continuous-time and renewal HSMMs [1611.01099]. Model selection—e.g., via Bayesian marginal likelihood/bridge sampling or AIC/BIC—can reliably distinguish duration families and allocate necessary state complexity [2006.09061, 1203.3485, 2405.13553]. ## 7. Extensions and Practical Modeling Considerations Numerous structural and algorithmic innovations have been proposed within the HSMM framework: - Nonhomogeneous/covariate-dependent dwell modeling [2312.14719, 2109.09949, 2405.13553], - Interval modeling and state gaps [1608.06954], - Bayesian nonparametrics/HDPMMs [1203.3485], - Higher-order transition chains and latent class decompositions [2002.05272], - Online/streaming algorithms and sequential Bayesian estimation [2301.10494], - Hybrid Markov/semi-Markov topologies for complex state networks [2109.12489], - Nonparametric emission modeling (e.g., B-splines) and penalized regressions [2109.12489]. These advances enable HSMMs to accommodate absorbing/macro-states, irregular sampling, informative censoring, and regime switching in domains ranging from environmental forecasting and robotics to health prognostics and energy systems. --- **References** - [2312.14719]: Nonhomogeneous hidden semi-Markov models for toroidal data - [1608.06954]: State Duration and Interval Modeling in Hidden Semi-Markov Model for Sequential Data Analysis - [2302.05347]: Bayesian Sparse Vector Autoregressive Switching Models with Application to Human Gesture Phase Segmentation - [2109.09949], [2010.10739]: Covariate-dependent duration HSMMs in high-frequency data - [1203.3485]: The Hierarchical Dirichlet Process Hidden Semi-Markov Model - [2006.09061]: Bayesian Approximations to Hidden Semi-Markov Models - [1407.3422]: A Spectral Algorithm for Inference in Hidden Semi-Markov Models - [2405.13553]: Inhomogeneous HSMMs with periodic dwell-time structure - [2210.09058]: Forward-Backward Latent State Inference for Hidden Continuous-Time semi-Markov Chains - [1611.01099]: Informational and Causal Architecture of Continuous-time Renewal and Hidden Semi-Markov Processes - [1811.07489]: Generalizing Robot Imitation Learning with Invariant Hidden Semi-Markov Models - [2002.05272]: Health Assessment and Prognostics Based on Higher Order Hidden Semi-Markov Models - [2109.12489]: Hhsmm: An R package for hidden hybrid Markov/semi-Markov models - [2301.10494]: Sequential Bayesian Learning for Hidden Semi-Markov Models - [1710.03914]: Backward Approximate Dynamic Programming with Hidden Semi-Markov Stochastic Models in Energy Storage Optimization - [2105.11490]: Hidden Markov and semi-Markov models: When and why are these models useful for classifying states in time series data Source: https://www.emergentmind.com/topics/hidden-semi-markov-models-hsmms