CDC2000 Estimation and Identification using Hidden Markov Models

Estimation and Identification using Hidden Markov Models
Home Full List of Titles 1: Proceedings of CDC2000 Discrete Event Systems Control in Communication Systems Optimal Control and Applications I Optimisation Approaches and Methods Model Predictive Control Advances in Linear Estimation Stochastic and Uncertain Systems Nonlinear Control and Applications Nonlinear Estimation and Filtering Formation Control and its Applications New Approaches to Fuzzy Control Manufacturing Systems Automotive Applications Stability Issues in Hybrid Control Recent Advances in Stochastic Networks Optimal Control and Applications II Robust Controller Design - mu, L1 and H2 Constrained and Receding Horizon Control Identification and Control around the World Markov Decision Processes Nonlinear Optimisation Observers for Nonlinear Systems Motion Planning Neural / Fuzzy Stability and Control Motor Control Control of Quantum Phenomena I Hybrid Systems Methods Control in Communication Networks Robustness and Optimisation Bumpless Transfer, Antiwindup and Saturation Adaptive Control: Linear Systems Estimation and Closed Loop Identification Control of Markov Processes Nonlinear Filtering and Control Modelling, Identification and Validation of Nonlinear Systems Differential Geometric Control Theory for Mechanical Systems Nonlinear Output Feedback Control Pneumatics and Compression Systems Control of Quantum Phenomena II Stability of Hybrid Systems Performance Analysis in Communication Networks Adaptive Control of Nonlinear Systems LMI Methods in Design Robust Control of Time Delay Systems Subspace Identification Methods Nonlinear Stochastic Filtering and Estimation Bifurcations, Chaos and Control I New Progress in Synthesis of Nonlinear Systems I Implementation Issues of Sliding Mode Control Theory Control of Mixing in Shear Flows Novel Neural Network Control Techniques for Industrial Motion Control Systems Physiological Control Systems Optimal Control of Hybrid Systems Stochastic Models for Communication Networks Control and Stabilisation of Nonlinear Systems New Directions in Robust Control Linear Systems Theory Advanced Topics in Systems Theory Estimation in Action Bifurcations, Chaos and Control II New Progress in Synthesis of Nonlinear Systems II Numerical Design and Analysis Techniques for Nonlinear Systems Analysis and Control of Underactuated Systems Sliding Mode Control I Challenges in the Application of Control to Computer Systems Estimation and Diagnosis of Discrete Event Systems Communications and Games Optimal Control Stochastic Systems Model Reduction Methodologies Identification and Subspace Methods Applications of Nonlinear Adaptive Control Advances in Nonlinear Output Feedback Design The Behavioural Approach to Systems and Control Vision Based Estimation and Control: Recent Advances and Open Problems Agile Control of Military Operations Sliding Mode Control II Model-based Fault Diagnosis of Industrial Processes Discrete Event Systems / Petri Nets System Identification and Confidence Estimation New Approaches to H-Infinity Control I Probabilistic Approaches to Robust Control Time Delay System Stabilisation Identification Methods Controlled Stochastic Processes Output Feedback of Nonlinear Systems Topics in Nonlinear Stabilisation Mobile Robots: Tracking Control Robust Control of Nonlinear Systems Power Systems Stabilisation and Control Disk Drive Control Hybrid Control Applications Discrete Time Systems New Approaches to H-Infinity Control II Linear Systems with Saturating Actuators New Theories in Distributed Parameter Systems Applications of Estimation and Identification Stochastic Control and Tuning Methodologies Control of Nonlinear Systems Iterative Learning and Control Coordinating Robot Systems Nonlinear Time Varying Systems Novel Applications of Neural Networks Aerospace Applications Switched Systems Implicit and Descriptor Systems LQG Periodic Systems and Disturbances New Horizons for Distributed Parameter Systems State Estimation Learning and Neuro-Control Nonlinear Control and Stabilisation I Tracking Vision Servoing Controllability of Nonlinear Systems Control of Flexible Systems Electro-Mechanical Systems Robust Control Methods and Applications Fault Detection and Diagnosis Optimisation and Applications Robust Stability Analysis Numerical Methods in Control Filtering in Continuous Time Stochastic Systems Interplay between Control and Signal Processing Fault Detection and Analysis Nonlinear Dynamical Systems Nonlinear Time Delay Systems Computational Issues in Nonlinear Control Disturbance Rejection Process Control Industry Applications Linear Parameter Varying Systems Linear Control Systems Dynamic and Nonlinear Programming Model Reduction Applications New Techniques for Control and Systems: Numerical Linear Algebra Estimation and Identification using Hidden Markov Models Applications of Stochastic Control Topics in Linear Design Nonlinear Control and Stabilisation II Ambulatory Robot Systems Chaotic and Oscillatory Systems Biomedical System Control Integrated Control and CPU Scheduling Linear Design Techniques Adaptive Disturbance / Noise Compensation Nonlinear Model Predictive Control Sensitivity Design, Analysis and Limitations Analysis of Linear Systems Linear Matrix Inequalities in Design Lyapunov's 2nd Method Robotics: Tracking Control Lagrangian and Hamiltonian Theory Variable Structure Control Machine Vision Signal Processing Methods in Control Applied Nonlinear Control Author Index A B C D E F G H I J K L M N O P Q R S T U V W X Y Z	Robust EM Algorithms For Markov Modulated Poisson Processes Authors: Robert J. Elliott, William P. Malcolm, Volume: 1, Page 4678 Paper number 1419 Abstract: In this article we consider robust filtering and smoothing for Markov Modulated Poisson Processes (MMPPs). Using the EM algorithm, these filters and smoothers can be applied to estimate the parameters of our model. Our dynamics do not involve stochastic integrals and our new formulae, in terms of time integrals, are easily discretized. CD001419.PDF (From Author) TOP Fault Detection in Hidden Markov Models: A Local Asymptotic Approach Authors: François LeGland, Laurent Mével, Volume: 1, Page 4686 Paper number 1829 Abstract: In this paper, the problem of detecting a change in the transition probability matrix of a hidden Markov chain is addressed, using the local asymptotic approach. The score function, evaluated at the nominal value, is used as the residual, and is expressed as an additive functional of the extended Markov chain consisting of the hidden state, the observation, the prediction filter and its gradient w.r.t. the parameter. The problem of residual evaluation is solved using available limit theorems on the extended Markov chain, which allow to replace the original detection problem by the simpler problem of detecting a change in the mean of a Gaussian r.v. CD001829.PDF (From Author) TOP Convergence Rates of the Maximum Likelihood Estimator of Hidden Markov Models Authors: Laurent Mével, Lorenzo Finesso, Volume: 1, Page 4691 Paper number 1664 Abstract: In this paper we consider the problem of the identification of a partially observed finite state Markov chain (or Hidden Markov Model, HMM), with continuous observations. Maximum Likelihood (ML) is the most popular approach to parameter estimation for this class of models. The asymptotic properties of the ML estimator (MLE) have already been investigated under a variety of conditions. Under the assumption of stationarity, Leroux has proved the almost sure consistency of the MLE, and Bykel, Ritov and Ryden have proved its asymptotic normality. More general results, encompassing both previous ones, have been given by Mevel. A new technique for the study of the convergence of HMM's has been developed. This technique is based on geometric ergodicity properties of the prediction filter and its derivatives, derived via results for products of random matrices. The main advantage is that convergence results for Hidden Markov Models can now be reduced to the analysis of a Markov process, with a properly defined state space. In this paper we apply the new technique to derive the almost sure rate of convergence of the MLE and give an example of application in the context of the model selection problem. As will be mentioned in the paper, these results extend easily to conditional least squares estimators. CD001664.PDF (From Author) TOP Reduced-Complexity Smoothing for Hidden Markov Models Authors: Louis Shue, Subhrakanti Dey, Volume: 1, Page 4697 Paper number 1167 Abstract: In this paper, we investigate approximate smoothing schemes for a class of hidden Markov models (HMM), namely, HMMs with underlying Markov chains that are nearly completely decomposable. The objective is to obtain substantial computational savings. Our algorithm can not only be used to obtain aggregate smoothed estimates, but can be used to obtain systematically approximate full-order smoothed estimates with computational savings, unlike many of the aggregation methods proposed earlier. CD001167.PDF (From Author) TOP On-Line Identification of Non-Minimum Phase Finite Impulse Response Linear Systems with Discrete Inputs via Hidden Markov Models Authors: Jason J. Ford, Volume: 1, Page 4703 Paper number 1032 Abstract: This paper considers adaptive estimation of noisy finite impulse response linear systems driven by stationary inputs from a discrete set. An equivalent hidden Markov model representations is presented which allows some powerful signal processing techniques to be applied to this estimation problem. A new adaptive estimation algorithm is presented and simulation studies illustrate the performance of the proposed algorithm. CD001032.PDF (From Author) TOP