On stability of nonlinear AR processes with Markov switching
The main purpose of this paper is to investigate the stability problem for the following nonlinear autoregressive (NAR) model with Markov switching (NAR-MS): Let \({\mathbf X}= (X_n)\) be a positive Markov chain on a finite set \(\{1,2,\dots, m\}\), and a NAR-MS process \({\mathbf Y}\) defined by the recurrent relation \[ Y_n= f_{X_n}(Y_{n- 1})+ \varepsilon_n,\quad Y_n\in{\mathbf R}^d,\quad n\geq 1.\tag{YM} \] \(\varepsilon= (\varepsilon_n)_{n\geq 0}\) is the error process, \((f_k)\), \(k= 1,2,\dots, m\), is a family of NAR functions, and \(\varepsilon\), \({\mathbf X}\), and \(Y_0\) are assumed to be independent. The stability of the NAR-MS model \({\mathbf Y}\) is manifested by obtaining suitable conditions to solve the following three problems: Problem A: The existence and the uniqueness of a strictly stationary ergodic solution for the NAR-MS model \({\mathbf Y}\); Problem B: The existence of the moments for the solution of \({\mathbf Y}\) considered in Problem A; Problem C: To establish limit theorems (including strong law of large numbers and the central limit theorem) as a basic tool for an asymptotic estimation theory of the NAR-MS model \({\mathbf Y}\). Section 2 introduces an extended Markov chain associated with the NAR-MS model \({\mathbf Y}\), and establishes some preliminary results. Section 3 investigates the Problems A, B, and C and provides solutions for the case when the autoregressive functions \(f_k\) \((k=1,2,\dots, m)\) in (YM) are sublinear functions. Section 4 analyses the same three problems for the case when the functions \(f_k\) are assumed to be Lipschitz ones.
- Stabilité des modèles AR fonctionnels à régime markovien
- Probabilistic Properties of a Nonlinear ARMA Process with Markov Switching
- Asymptotic properties of nonlinear autoregressive Markov processes with state-dependent switching
- A proof of consistency of the MLE for nonlinear Markov-switching AR processes
- Stochastic stability for the nonlinear autoregressive series
- On the stationary law of a nonlinear autoregressive Markov chain
- On least squares estimation for stable nonlinear AR processes
- Asymptotic properties of the maximum likelihood estimator in autoregressive models with Markov regime
- Stability of a random diffusion with nonlinear drift
- Misspecified structural change, threshold, and Markov-switching models.
- Penalized estimate of the number of states in Gaussian linear AR with Markov regime
- Exponential forgetting of smoothing distributions for pairwise Markov models
- Stability analysis of switched ARX models and application to learning with guarantees
- Consistency of quasi-maximum likelihood estimator for Markov-switching bilinear time series models
- Ergodicity and existence of moments for local mixtures of linear autoregressions
- Blockwise bootstrap testing for stationarity
- Some theoretical results on Markov-switching autoregressive models with gamma innovations
- A proof of consistency of the MLE for nonlinear Markov-switching AR processes
- Consistency of the maximum likelihood estimate for non-homogeneous Markov-switching models
- On the Markov-switching bilinear processes: stationarity, higher-order moments and -mixing
- On Markov-switching ARMA processes-stationarity, existence of moments, and geometric ergodicity
- Stabilité des modèles AR fonctionnels à régime markovien
- Linear diffusion with stationary switching regime
- scientific article; zbMATH DE number 2065863 (Why is no real title available?)
- Probabilistic Properties of a Nonlinear ARMA Process with Markov Switching
- Hidden Markov mixture autoregressive models: stability and moments
- Sparseness, consistency and model selection for Markov regime-switching Gaussian autoregressive models
- First-order integer-valued autoregressive process with Markov-switching coefficients
- Linear autoregressive models
- Asymptotic analysis of temporal-difference learning algorithms with constant step-sizes
- Asymptotic analysis of temporal-difference learning algorithms with constant step-sizes
- Stationarity and ergodic properties for some observation-driven models in random environments
- Spectral representation of Markov-switching bilinear processes
- Covariance analysis and GMM estimation of Markov switching bilinear processes
- A general autoregressive model with Markov switching: estimation and consistency
- Tail behaviour and extremes of two-state Markov-switching autoregressive models
- Asymptotic properties of nonlinear autoregressive Markov processes with state-dependent switching
This page was built for publication: On stability of nonlinear AR processes with Markov switching
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4507946)
