Inference in Hidden Markov Models

By Haikady N. Nagaraja. 2006

1,470 citations recorded by the source corpus.

Subfields: Computer Networks and Communications, Probabilistic & Bayesian Methods

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Networks and Communications

In Probabilistic & Bayesian Methods

Citations in the corpus

Sequential Kernel Herding: Frank-Wolfe Optimization for Particle Filtering (2015, background): “This poses diﬃculties for particle-ﬁltering-based solutions for two reasons: (1) the computational bottleneck arising from the likelihood evaluation implies that we cannot simply increase the number of particles to improve the accuracy, and (2) this type of “complicated” observation models will…”
Identifiability of parameters in latent structure models with many observed variables (2008, methodology): “We refer to [5, 16] for more details on HMMs.”
The sample size required in importance sampling (2015, methodology): “We recommend Hammersley and Handscomb [29], Srinivasan [55], Cappé, Moulines and Rydén [13] and Liu [40].”
An Overview of Existing Methods and Recent Advances in Sequential Monte Carlo (2007, methodology): “Note that this algorithm is genuinely sequential in that bx0:t approximates the MAP estimator for each and every t (3)It should be noted that the backward function is not a probability distribution (it is not, and in some cases it may not be, normalized), see [7], [72] and [73] for further discussio”
Discrete-Time Nonlinear Filtering Algorithms Using Gauss–Hermite Quadrature (2007, methodology): “The SMC method consists of a large number of variants; many more are being constantly proposed in an attempt to improve both the statistical and computational efficiency (see [12], [13], [16], and the references therein).”
Biips: Software for Bayesian Inference with Interacting Particle Systems (2014, methodology): “Traditionally, SMC methods have been restricted to the class of state-space models or hidden Markov chain models for which those models are particularly suited (Cappé et al. 2005).”
Adaptive sequential Monte Carlo by means of mixture of experts (2011, methodology): “Common choices are the widely used squared coefﬁcient of variation of the importance weights or the negated Shannon entropy (see e.g. Cappé et al. 2005, p. 235, for deﬁnitions of these quantities) of the same.”
Hidden Markov models with t components. Increased persistence and other aspects (2011, methodology): “A more comprehensive and detailed survey on HMMs was reported by Cappé et al. (2007).”
Space–time modelling of precipitation by using a hidden Markov model and censored Gaussian distributions (2009, methodology): “Various procedures have been proposed in the literature for the choice of N(n) (see [7], [12] and [6] for example) including both deterministic (e.”
SMC2: an efficient algorithm for sequential analysis of state space models (2011, methodology): “g the existence of sufficient statistics), an on-line algorithm may be obtained by extending SOPF so as to include MCMC updates of the θ-component, see Gilks and Berzuini (2001), Fearnhead (2002), Storvik (2002), and also the more recent work of Carvalho et al. (2010), but numerical evidence seems t”
Two‐part regression models for longitudinal zero‐inflated count data (2010, background): “Although they have been introduced in the late 1960s, hidden Markov models (HMMs, for details see MacDonald & Zucchini 1997; Cappé et al. 2005) have become increasingly popular in the last 10 years, due to their rich mathematical structure and flexibility.”
Hidden semi-Markov models (2010, neutral): “Detailed review for the conventional HMM can be found in the tutorial by Rabiner [150], the overview by Ephraim and Merhav [57], the Bayesian networks-based discussion by Ghahramani [66], and the book by Cappe et al. [29].”
Electricity price forecasting: A review of the state-of-the-art with a look into the future (2014, neutral): “Jonsson et al. (2014) generate prediction densities of day-ahead”
Variational Sequential Monte Carlo (2017, background): “Two concrete examples, both from the time-series literature, are hidden Markov models and state space models [Cappé et al., 2005].”
Some Parameter Estimation Issues in Functional-Structural Plant Modelling (2011, background): “[6]): model equations can be interpreted as a bidimensional stochastic process, which describes the evolution of system variables.”
Unsupervised Machine Learning for Networking: Techniques, Applications and Research Challenges (2017, background): “It is for this reason that an HMM is defined to be a doubly stochastic process: first, the underlying stochastic process is not observable; and second, another stochastic process, dependent on the underlying stochastic process, that produces a sequence of observed symbols [93].”
Inference in finite state space non parametric Hidden Markov Models and applications (2015, methodology): “Efﬁcient algorithms allow to compute the likelihood and to build practical inference methods for both mixture and HMMs, see e.g. Cappé et al. (2005) for a recent state of the art in HMMs.”