Graphical model selection for a particular class of continuous-time processes

Zorzi, Mattia

Kybernetika, Tome 55 (2019) no. 5, pp. 782-801

Abstract (VO)

Graphical models provide an undirected graph representation of relations between the components of a random vector. In the Gaussian case such an undirected graph is used to describe conditional independence relations among such components. In this paper, we consider a continuous-time Gaussian model which is accessible to observations only at time $T$. We introduce the concept of infinitesimal conditional independence for such a model. Then, we address the corresponding graphical model selection problem, i. e. the problem to estimate the graphical model from data. Finally, simulation studies are proposed to test the effectiveness of the graphical model selection procedure.

Détail

DOI : 10.14736/kyb-2019-5-0782

Classification : 65K10, 93B30
Keywords: sparse inverse covariance selection; regularization; graphical models; entropy; optimization

Bibliographie

[1] Alpago, D., Zorzi, M., Ferrante, A.: Identification of sparse reciprocal graphical models. IEEE Control Systems Lett. 2 (2018), 4, 659-664. | DOI

[2] Avventi, E., Lindquist, A., Wahlberg, B.: ARMA identification of graphical models. IEEE Trans. Automat. Control 58 (2013), 1167-1178. | DOI | MR

[3] Baggio, G.: Further results on the convergence of the Pavon-Ferrante algorithm for spectral estimation. IEEE Trans. Automat- Control 63 (2018), 10, 3510-3515. | DOI | MR

[4] Banerjee, O., Ghaoui, L. El, d'Aspremont, A.: Model selection through sparse maximum likelihood estimation for multivariate Gaussian or binary data. J. Machine Learning Res. 9 (2008), 485-516. | MR

[5] Boyd, S., Vandenberghe, L.: Convex Optimization. Cambridge Univ. Press, Cambridge 2004. | DOI | MR | Zbl

[6] Byrnes, C., Gusev, S., Lindquist, A.: A convex optimization approach to the rational covariance extension problem. SIAM J. Optim. 37 (1998), 211-229. | DOI | MR

[7] Byrnes, C. I., Georgiou, T. T., Lindquist, A.: A new approach to spectral estimation: A tunable high-resolution spectral estimator. IEEE Trans. Signal Process. 48 (2000), 3189-3205. | DOI | MR

[8] Candes, E., Plan, Y.: Matrix completion with noise. Proc. IEEE 98 (2010), 925-936. | DOI

[9] Candes, E., Recht, B.: Exact matrix completion via convex optimization. Comm. ACM 55 (2012), 111-119. | DOI | MR

[10] Chandrasekaran, V., Parrilo, P., Willsky, A.: Latent variable graphical model selection via convex optimization. Ann. Statist. 40 (2010), 1935-2013. | DOI | MR

[11] Chandrasekaran, V., Shah, P.: Relative entropy optimization and its applications. Math. Program. 161 (2017), (1-2), 1-32. | DOI | MR

[12] Cover, T., Thomas, J.: Information Theory. Wiley, New York 1991. | DOI

[13] d'Aspremont, A., Banerjee, O., Ghaoui, L. El: First-order methods for sparse covariance selection. SIAM J. Matrix Analysis Appl. 30 (2008), 56-66. | DOI | MR

[14] Dempster, A.: Covariance selection. Biometrics 28 (1972), 157-175. | DOI | MR

[15] Ferrante, A., Pavon, M.: Matrix completion à la Dempster by the principle of parsimony. IEEE Trans. Inform. Theory 57 (2011), 3925-3931. | DOI | MR

[16] Ferrante, A., Pavon, M., Ramponi, F.: Hellinger versus Kullback-Leibler multivariable spectrum approximation. IEEE Trans. Autom. Control 53 (2008), 954-967. | DOI | MR

[17] Friedman, J., Hastie, T., Tibshirani, R.: Sparse inverse covariance estimation with the graphical lasso. Biostatistics 9 (2008), 432-441. | DOI

[18] Grant, M., Boyd, S.: CVX: Matlab software for disciplined convex programming, version 2.1. 2014.

[19] Gu, S., Betzel, R., Mattar, M., Cieslak, M., Delio, P., Grafton, S., Pasqualetti, F., Bassett, D.: Optimal trajectories of brain state transitions. NeuroImage 148 (2017), 305-317. | DOI

[20] Huang, J., Liu, N., Pourahmadi, M., Liu, L.: Covariance matrix selection and estimation via penalised normal likelihood. Biometrika 93 (2006), 85-98. | DOI | MR

[21] Huotari, N., Raitamaa, L., Helakari, H., Kananen, J., Raatikainen, V., Rasila, A., Tuovinen, T., Kantola, J., Borchardt, V., Kiviniemi, V., Korhonen, V.: Sampling rate effects on resting state fMRI metrics. Frontiers Neurosci. 13 (2019), 279. | DOI

[22] Jalali, A., Sanghavi, S.: Learning the dependence graph of time series with latent factors. In: International Conference on Machine Learning Edinburgh 2012.

[23] Koller, D., Friedman, N.: Probabilistic Graphical Models: Principles and Techniques. MIT Press, 2009. | MR

[24] Lauritzen, S.: Graphical Models. Oxford University Press, Oxford 1996. | MR

[25] Meinshausen, N., Bühlmann, P.: High-dimensional graphs and variable selection with the lasso. Annals Statist. 34 (2006), 1436-1462. | DOI | MR

[26] Pearl, J.: Graphical models for probabilistic and causal reasoning. In: Quantified representation of uncertainty and imprecision, Springer 1998, pp. 367-389. | DOI | MR

[27] Ringh, A., Karlsson, J., Lindquist, A.: Multidimensional rational covariance extension with approximate covariance matching. SIAM J. Control Optim. 56 (2018), 2, 913-944. | DOI | MR

[28] Songsiri, J., Dahl, J., Vandenberghe, L.: Graphical models of autoregressive processes. In: Convex Optimization in Signal Processing and Communications (D. Palomar and Y. Eldar, eds.), Cambridge Univ. Press, Cambridge 2010, pp. 1-29. | MR

[29] Songsiri, J., Vandenberghe, L.: Topology selection in graphical models of autoregressive processes. J. Machine Learning Res. 11 (2010), 2671-2705. | MR

[30] Yue, Z., Thunberg, J., Ljung, L., Gonçalves, J.: Identification of sparse continuous-time linear systems with low sampling rate: Exploring matrix logarithms. arXiv preprint arXiv:1605.08590, 2016.

[31] {Zhu}, B., {Baggio}, G.: On the existence of a solution to a spectral estimation problem a la Byrnes-Georgiou-Lindquist. IEEE Trans. Automat. Control 64 (2019), 2, 820-825. | DOI | MR

[32] Zorzi, M.: A new family of high-resolution multivariate spectral estimators. IEEE Trans. Automat. Control 59 (2014), 892-904. | DOI | MR

[33] Zorzi, M.: Rational approximations of spectral densities based on the Alpha divergence. Math. Control Signals Systems 26 (2014), 259-278. | DOI | MR

[34] Zorzi, M.: An interpretation of the dual problem of the THREE-like approaches. Automatica 62 (2015), 87-92. | DOI | MR

[35] Zorzi, M.: Multivariate Spectral Estimation based on the concept of Optimal Prediction. IEEE Trans. Automat. Control 60 (2015), 1647-1652. | DOI | MR

[36] Zorzi, M.: Empirical Bayesian learning in AR graphical models. Automatica 109 (2019), 108516. | DOI | MR

[37] Zorzi, M., Sepulchre, R.: AR identification of latent-variable graphical models. IEEE Trans. Automat. Control 61 (2016), 2327-2340. | DOI | MR