An Empirical Evaluation of Robust Gaussian Process Models for System Identification
System identification comprises a number of linear and nonlinear tools for black-box modeling of dynamical systems, with applications in several areas of engineering, control, biology and economy. However, the usual Gaussian noise assumption is not always satisfied, specially if data is corrupted by impulsive noise or outliers. Bearing this in mind, the present paper aims at evaluating how Gaussian Process (GP) models perform in system identification tasks in the presence of outliers. More specifically, we compare the performances of two existing robust GP-based regression models in experiments involving five benchmarking datasets with controlled outlier inclusion. The results indicate that, although still sensitive in some degree to the presence of outliers, the robust models are indeed able to achieve lower prediction errors in corrupted scenarios when compared to conventional GP-based approach.
KeywordsRobust system identification Gaussian process Approximate Bayesian inference
The authors thank the financial support of FUNCAP, IFCE, NUTEC and CNPq (grant no. 309841/2012-7).
- 3.Lawrence, N.D.: Gaussian process latent variable models for visualisation of high dimensional data. In: Advances in Neural Information Processing Systems, pp. 329–336 (2004)Google Scholar
- 5.Murray-Smith, R., Johansen, T.A., Shorten, R.: On transient dynamics, off-equilibrium behaviour and identification in blended multiple model structures. In: European Control Conference (ECC 1999), Karlsruhe, BA-14. Springer (1999)Google Scholar
- 6.Solak, E., Murray-Smith, R., Leithead, W.E., Leith, D.J., Rasmussen, C.E.: Derivative observations in Gaussian process models of dynamic systems. In: Advances in Neural Information Processing Systems 16 (2003)Google Scholar
- 10.Frigola, R., Chen, Y., Rasmussen, C.: Variational Gaussian process state-space models. In: Advances in Neural Information Processing Systems (NIPS), vol. 27, pp. 3680–3688 (2014)Google Scholar
- 12.Minka, T.P.: Expectation propagation for approximate Bayesian inference. In: Proceedings of the 17th Conference on Uncertainty in Artificial Intelligence (UAI 2001), pp. 362–369. Morgan Kaufmann (2001)Google Scholar
- 14.Kuss, M., Pfingsten, T., Csató, L., Rasmussen, C.E.: Approximate inference for robust Gaussian process regression. Technical report 136, Max Planck Institute for Biological Cybernetics, Tubingen, Germany (2005)Google Scholar
- 17.Berger, B., Rauscher, F.: Robust Gaussian process modelling for engine calibration. In: Proceedings of the 7th Vienna International Conference on Mathematical Modelling (MATHMOD 2012), pp. 159–164 (2012)Google Scholar
- 18.Kuss, M.: Gaussian process models for robust regression, classification, and reinforcement learning. Ph.D. thesis, TU Darmstadt (2006)Google Scholar
- 19.Rasmussen, C.E.: Evaluation of Gaussian processes and other methods for non-linear regression. Ph.D. thesis, University of Toronto, Toronto, Canada (1996)Google Scholar