Diagnostic Checking in a Flexible Nonlinear Time Series Model
This paper deals with a sequence of misspecification tests for a flexible nonlinear time series model, called the neuro-coefficient smooth transition autoregressive (NCSTAR) model. The flexible nonlinear NCSTAR model has the form \[ y_{t}=\alpha'z_{t}'+ \sum_{i=1}^{h}\lambda_{i}' z_{t}F(\omega_{i}'x_{t}-\beta_{i})+ \varepsilon_{t}. \] The vector \(z_{t}\) is defined as \(z_{t}=[1,\widetilde z_{t}']'\), where \(\widetilde z_{t}\) is a vector of lagged values of \(y_{t}\) and some exogenous variables. The function \(F(\omega_{i}'x_{t}-\beta_{i})\) is the logistic function, where \(x_{t}\) is a vector of transition variables, \(\omega_{i}\) and \(\beta_{i}\) are real parameters, and \({\varepsilon_{t}}\) is a sequence of independent, normally distributed random variables with mean zero. The authors consider Lagrange multiplier type tests for testing the hypotheses of parameter constancy, serial independence, and homoscedasticity. A Monte Carlo simulation shows that the tests are well sized and have good power in small samples.
- Testing the adequacy of smooth transition autoregressive models
- Specification, estimation, and evaluation of smooth transition autoregressive models
- A Portmanteau Test for Smooth Transition Autoregressive Models
- Testing linearity against smooth transition autoregressive models
- SMOOTH TRANSITION AUTOREGRESSIVE MODELS — A SURVEY OF RECENT DEVELOPMENTS
- Diagnostic checking in linear processes with infinite variance
- Asymptotic theory for regressions with smoothly changing parameters
- Diagnostic check for heavy tail in linear time series
- Smooth transition autoregressive models and fuzzy rule-based systems: Functional equivalence and consequences
- Theory study and empirical analysis on nonlinearity tests for an LSTAR model with two thresholds
- A comparison of statistical tests for the adequacy of a neural network regression model
- A multiple regime smooth transition heterogeneous autoregressive model for long memory and asymmetries
- Testing for remaining autocorrelation of the residuals in the framework of fuzzy rule-based time series modelling
- Testing constancy of the error covariance matrix in vector models
- On detecting the optimal structure of a neural network under strong statistical features in errors
This page was built for publication: Diagnostic Checking in a Flexible Nonlinear Time Series Model
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4455661)
