Recursive estimation in ARMAX models
\(M\)-estimatorsadaptive controlARMAX systemasymptotic efficiencyasymptotic normalityautoregressive moving average with exogenous inputsbackshift operatorconsistencyconsistent recursive estimatorsextended least squaresextended stochastic Lyapunov functionslinear stochastic difference equationmartingale theorymethod of momentsmonitored maximum likelihoodon-line predictionrecursive least squares
[For the entire collection see Zbl 0761.00004.] The ARMAX system (autoregressive moving average with exogenous inputs) is defined by the linear stochastic difference equation \[ A(q^{-1})y_ n=q^{-d} B(q^{-1}) u_ n+C(q^{-1}) \varepsilon_ n, \] where \(\{u_ n\}\), \(\{y_ n\}\), \(\{\varepsilon_ n\}\) denote the input, output and disturbance sequences, \(d\geq 1\) is the delay, and \(A\), \(B\), \(C\) are scalar polynomials in the backshift operator \((q^{-1} x_ n=x_{n-1})\). In sections 2 and 3 the author reviews two classes of recursive estimators treated in the literature, namely, the stochastic approximation approach and associated martingale theory, and recursive and extended least squares. In these sections the author also outlines some of the methods used to analyze these algorithms, in particular the method of stochastic and extended stochastic Lyapunov functions. These ideas are then related, in two further sections, to recent developments on consistent recursive estimators based on the method of moments, and monitored recursive maximum likelihood and \(M\)-estimators. ``These developments not only extend important statistical properties such as consistency, asymptotic normality, asymptotic efficiency,... but are also applicable to on-line prediction and adaptive control of ARMAX systems.
- Recursive Identification and Adaptive Prediction in Linear Stochastic Systems
- Suboptimum Maximum Likelihood Identification of ARMAX Processes
- scientific article; zbMATH DE number 440435
- On the generalized least squares estimator of an ARMAX model. Application to the identification of ARMA models
- Efficient recursive estimation and adaptive control in stochastic regression and ARMAX mod\-els
- Estimating ARMAX systems for multivariate time series using the state approach to subspace algorithms
- Stochastic adaptive pole-zero assignment with convergence analysis
- On the generalized least squares estimator of an ARMAX model. Application to the identification of ARMA models
- Optimal calculation of residuals for ARMAX models with application to model verification
- Recursive method for ARMA model estimation. I
- Recursive max-linear models with propagating noise
- A new algorithm for recursive estimation of parameters in controlled ARMA processes
- On identifiability for multidimensional ARMAX model
- Estimation of nonstationary ARMAX models based on the Hannan-Rissanen method
- Recursive parameter method for computing the predicting function of the multivariable ARMAX model
- scientific article; zbMATH DE number 440435 (Why is no real title available?)
- Efficient recursive estimation and adaptive control in stochastic regression and ARMAX mod\-els
- scientific article; zbMATH DE number 4106081 (Why is no real title available?)
- Choice of least-squares algorithms for the identification of ARM AX dynamic systems
- Suboptimum Maximum Likelihood Identification of ARMAX Processes
- Constructing NARMAX models using ARMAX models
- ARMAX Modelling in Structural Dynamics – Numerical Integration of the Integral Representations
- Risk sensitive identification of ARMA processes
- A recursive algorithm for ARMAX model identification in closed loop
- Estimation and prediction with ARMMAX model: a mixture of ARMAX models with common ARX part
- Weighted Estimation and Tracking for ARMAX Models
- scientific article; zbMATH DE number 5206499 (Why is no real title available?)
- A Note on the Specification and Estimation of ARMAX Systems
- Projection-based Bayesian recursive estimation of ARX model with uniform innovations
This page was built for publication: Recursive estimation in ARMAX models
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4694364)
