Latent Autoregressive Gaussian Processes Models for Robust System Identification

Latent Autoregressive Gaussian Process Models for Robust System Identification

We introduce GP-RLARX, a novel Gaussian Process (GP) model for robust system identification. Our approach draws inspiration from nonlinear autoregressive modeling with exogenous inputs (NARX) and it encapsulates a novel and powerful structure referred to as latent autoregression. This structure accounts for the feedback of uncertain values during training and provides a natural framework for fr...

Gaussian Processes for Functional-Coefficient Autoregressive Models

This work is concerned with nonlinear time series models and, in particular, with nonparametric models for the dynamics of the mean of the time series. We build on the functional-coefficient autoregressive (FAR) model of Chen and Tsay (1993) which is a generalization of the autoregressive (AR) model where the coefficients are varying and are given by functions of the lagged values of the series...

Inverse Gaussian Autoregressive Models

A first-order autoregressive process with inverse gaussian marginals is introduced. The innovation distributions are obtained under certain special cases. The unknown parameters are estimated using different methods and these estimators are shown to be consistent and asymptotically normal. The behavior of the estimators for small samples is studied through simulation experiments. On Sums of Tri...

Parameter estimation for non-Gaussian autoregressive processes

It is proposed to jointly estimate the parameters of nonGaussian autoregressive (AR) processes in a Bayesian context using the Gibbs sampler. Using the Markov chains produced by the sampler an approximation to the vector MAP estimator is implemented. The results reported here used AR(4) models driven by noise sequences where each sample is iid as a two component Gaussian sum mixture. The result...

Maximum Likelihood Identification of Gaussian Autoregressive Moving Average Models