Approximate Recursive Identification of Autoregressive Systems with Skewed Innovations

We propose a novel recursive system identification algorithm for linear autoregressive systems with skewed innovations. The algorithm is based on the variational Bayes approximation of the model with a multivariate normal prior for the model coefficients, multivariate skew-normally distributed innovations, and matrix-variate-normal - inverse-Wishart prior for the parameters of the innovation distribution. The proposed algorithm simultaneously estimates the model coefficients as well as the parameters of the innovation distribution, which are both allowed to be slowly time-varying. Through computer simulations, we compare the proposed method with a variational algorithm based on the normally-distributed innovations model, and show that modelling the skewness can provide improvement in identification accuracy.

[1]  Geoffrey J. McLachlan,et al.  Finite mixtures of canonical fundamental skew $$t$$t-distributions , 2014, Stat. Comput..

[2]  Eric Jondeau,et al.  Conditional volatility, skewness, and kurtosis: existence, persistence, and comovements , 2003 .

[3]  Eduardo Mario Nebot,et al.  Approximate Inference in State-Space Models With Heavy-Tailed Noise , 2012, IEEE Transactions on Signal Processing.

[4]  Peter Christoffersen,et al.  Série Scientifique Scientific Series Option Valuation with Conditional Skewness Option Valuation with Conditional Skewness , 2022 .

[5]  Dan Simon,et al.  Constrained Kalman filtering via density function truncation for turbofan engine health estimation , 2010, Int. J. Syst. Sci..

[6]  Simo Ali-Löytty,et al.  Kalman-type positioning filters with floor plan information , 2008, MoMM.

[7]  A. Azzalini.  A class of distributions which includes the normal ones , 1985 .

[8]  Henrik Ohlsson,et al.  On the estimation of transfer functions, regularizations and Gaussian processes - Revisited , 2012, Autom..

[9]  M. Kárný,et al.  Approximate Bayesian Recursive Estimation On Approximation Errors No . 2317 January 26 , 2012 , 2012 .

[10]  Radford M. Neal.  Pattern Recognition and Machine Learning , 2007, Technometrics.

[11]  Fredrik Gustafsson,et al.  A NLOS-robust TOA positioning filter based on a skew-t measurement noise model , 2015, 2015 International Conference on Indoor Positioning and Indoor Navigation (IPIN).

[12]  Campbell R. Harvey,et al.  Current Version : April 6 , 1999 Autoregressive Conditional Skewness , 1999 .

[13]  Lennart Ljung,et al.  Recursive identification algorithms , 2002 .

[14]  M. Genton,et al.  On fundamental skew distributions , 2005 .

[15]  Prashant Krishnamurthy,et al.  Analysis of WLAN's received signal strength indication for indoor location fingerprinting , 2012, Pervasive Mob. Comput..

[16]  Fredrik Gustafsson,et al.  Skew-$t$ Filter and Smoother With Improved Covariance Matrix Approximation , 2016, IEEE Transactions on Signal Processing.

[17]  A. Azzalini,et al.  The multivariate skew-normal distribution , 1996 .

[18]  Geoffrey J. McLachlan,et al.  Finite mixtures of multivariate skew t-distributions: some recent and new results , 2014, Stat. Comput..

[19]  Fredrik Gustafsson,et al.  Marginalized adaptive particle filtering for nonlinear models with unknown time-varying noise parameters , 2013, Autom..

[20]  A. Rukhin.  Matrix Variate Distributions , 1999, The Multivariate Normal Distribution.

[21]  Saumyadipta Pyne,et al.  Maximum likelihood inference for mixtures of skew Student-t-normal distributions through practical EM-type algorithms , 2012, Stat. Comput..

[22]  Giuseppe De Nicolao,et al.  A new kernel-based approach for linear system identification , 2010, Autom..

[23]  Tsung I. Lin,et al.  Maximum likelihood estimation for multivariate skew normal mixture models , 2009, J. Multivar. Anal..

[24]  D.G. Tzikas,et al.  The variational approximation for Bayesian inference , 2008, IEEE Signal Processing Magazine.

[25]  Fredrik Gustafsson,et al.  Robust Inference for State-Space Models with Skewed Measurement Noise , 2015, IEEE Signal Processing Letters.

[26]  Georgios Tsiotas.  On generalised asymmetric stochastic volatility models , 2012, Comput. Stat. Data Anal..