Change search
Link to record
Permanent link

Direct link
BETA
Alternative names
Publications (10 of 126) Show all publications
Ha, H., Welsh, J. S., Rojas, C. R. & Wahlberg, B. (2018). An analysis of the SPARSEVA estimate for the finite sample data case. Automatica, 96, 141-149
Open this publication in new window or tab >>An analysis of the SPARSEVA estimate for the finite sample data case
2018 (English)In: Automatica, ISSN 0005-1098, E-ISSN 1873-2836, Vol. 96, p. 141-149Article in journal (Refereed) Published
Abstract [en]

In this paper, we develop an upper bound for the SPARSEVA (SPARSe Estimation based on a VAlidation criterion) estimation error in a general scheme, i.e., when the cost function is strongly convex and the regularized norm is decomposable for a pair of subspaces. We show how this general bound can be applied to a sparse regression problem to obtain an upper bound of the estimation error for the traditional I-1 SPARSEVA problem. Numerical results are used to illustrate the effectiveness of the suggested bound. 

Place, publisher, year, edition, pages
PERGAMON-ELSEVIER SCIENCE LTD, 2018
Keywords
SPARSEVA estimate, Estimation error, Upper bound, Finite sample data
National Category
Computer and Information Sciences
Identifiers
urn:nbn:se:kth:diva-235563 (URN)10.1016/j.automatica.2018.06.046 (DOI)000444659500015 ()2-s2.0-850495675102-s2.0-85049567510 (Scopus ID)
Note

QC 20181001

Available from: 2018-10-01 Created: 2018-10-01 Last updated: 2018-10-01Bibliographically approved
Abdalmoaty, M. ., Rojas, C. R. & Hjalmarsson, H. (2018). Identication of a Class of Nonlinear Dynamical Networks. In: : . Paper presented at 18th IFAC Symposium on System Identification.
Open this publication in new window or tab >>Identication of a Class of Nonlinear Dynamical Networks
2018 (English)Conference paper, Published paper (Refereed)
Abstract [en]

Identifcation of dynamic networks has attracted considerable interest recently. So far the main focus has been on linear time-invariant networks. Meanwhile, most real-life systems exhibit nonlinear behaviors; consider, for example, two stochastic linear time-invariant systems connected in series, each of which has a nonlinearity at its output. The estimation problem in this case is recognized to be challenging, due to the analytical intractability of both the likelihood function and the optimal one-step ahead predictors of the measured nodes. In this contribution, we introduce a relatively simple prediction error method that may be used for the estimation of nonlinear dynamical networks. The estimator is defined using a deterministic predictor that is nonlinear in the known signals. The estimation problem can be defined using closed-form analytical expressions in several non-trivial cases, and Monte Carlo approximations are not necessarily required. We show, that this is the case for some block-oriented networks with no feedback loops and where all the nonlinear modules are polynomials. Consequently, the proposed method can be applied in situations considered challenging by current approaches. The performance of the estimation method is illustrated on a numerical simulation example.

Series
IFAC-PapersOnLine
Keywords
System Identication, Dynamical Networks, Stochastic Systems, Block-Oriented Models, Prediction Error Method.
National Category
Signal Processing Control Engineering
Research subject
Electrical Engineering; Electrical Engineering
Identifiers
urn:nbn:se:kth:diva-233639 (URN)
Conference
18th IFAC Symposium on System Identification
Funder
EU, European Research Council, 267381Swedish Research Council, 2015-05285Swedish Research Council, 2016-06079
Note

QC 20180829

Available from: 2018-08-27 Created: 2018-08-27 Last updated: 2018-08-29Bibliographically approved
Bjurgert, J., Valenzuela, P. E. & Rojas, C. R. (2018). On Adaptive Boosting for System Identification. IEEE Transactions on Neural Networks and Learning Systems, 29(9), 4510-4514
Open this publication in new window or tab >>On Adaptive Boosting for System Identification
2018 (English)In: IEEE Transactions on Neural Networks and Learning Systems, ISSN 2162-237X, E-ISSN 2162-2388, Vol. 29, no 9, p. 4510-4514Article in journal (Refereed) Published
Abstract [en]

In the field of machine learning, the algorithm Adaptive Boosting has been successfully applied to a wide range of regression and classification problems. However, to the best of the authors' knowledge, the use of this algorithm to estimate dynamical systems has not been exploited. In this brief, we explore the connection between Adaptive Boosting and system identification, and give examples of an identification method that makes use of this connection. We prove that the resulting estimate converges to the true underlying system for an output-error model structure under reasonable assumptions in the large sample limit and derive a bound of the model mismatch for the noise-free case.

Place, publisher, year, edition, pages
Institute of Electrical and Electronics Engineers (IEEE), 2018
Keywords
Adaptive algorithms, adaptive boosting, dynamical systems, orthonormal basis functions, system identification
National Category
Control Engineering
Identifiers
urn:nbn:se:kth:diva-235117 (URN)10.1109/TNNLS.2017.2754319 (DOI)000443083700049 ()29035231 (PubMedID)2-s2.0-85052677886 (Scopus ID)
Note

QC 20180919

Available from: 2018-09-19 Created: 2018-09-19 Last updated: 2018-09-19Bibliographically approved
Oomen, T. & Rojas, C. R. (2018). Sparse Iterative Learning Control (SPILC): When to Sample for Resource-Efficiency?. In: 2018 IEEE 15TH INTERNATIONAL WORKSHOP ON ADVANCED MOTION CONTROL (AMC): . Paper presented at 15th IEEE International Workshop on Advanced Motion Control (AMC), MAR 09-11, 2018, Shibaura Inst Technol, Toyosu Campus, Tokyo, JAPAN (pp. 497-502). IEEE
Open this publication in new window or tab >>Sparse Iterative Learning Control (SPILC): When to Sample for Resource-Efficiency?
2018 (English)In: 2018 IEEE 15TH INTERNATIONAL WORKSHOP ON ADVANCED MOTION CONTROL (AMC), IEEE, 2018, p. 497-502Conference paper, Published paper (Refereed)
Abstract [en]

Iterative learning control enables the determination of optimal command inputs by learning from measured data of previous tasks. The aim of this paper is to address the negative impact of trial-varying disturbances that contaminate these measurements, both in terms of resource-efficient implementations and performance degradation. The proposed method is an optimal framework for ILC that enforces sparsity and related structure on the command signal. This is achieved through a convex relaxation relying on l(1) regularization. The approach is demonstrated on a benchmark motion system, confirming substantial extensions compared to earlier results.

Place, publisher, year, edition, pages
IEEE, 2018
Series
International Workshop on Advanced Motion Control, ISSN 1943-6572
National Category
Electrical Engineering, Electronic Engineering, Information Engineering
Identifiers
urn:nbn:se:kth:diva-232932 (URN)000436348700078 ()2-s2.0-85048787229 (Scopus ID)978-1-5386-1946-9 (ISBN)
Conference
15th IEEE International Workshop on Advanced Motion Control (AMC), MAR 09-11, 2018, Shibaura Inst Technol, Toyosu Campus, Tokyo, JAPAN
Note

QC 20180807

Available from: 2018-08-07 Created: 2018-08-07 Last updated: 2018-08-07Bibliographically approved
Mueller, M. I., Valenzuela, P. E., Proutiere, A. & Rojas, C. R. (2017). A stochastic multi-armed bandit approach to nonparametric H-infinity-norm estimation. In: 2-s2.0-85046136421: . Paper presented at IEEE 56th Annual Conference on Decision and Control (CDC), DEC 12-15, 2017, Melbourne, AUSTRALIA (pp. 4632-4637). Institute of Electrical and Electronics Engineers (IEEE)
Open this publication in new window or tab >>A stochastic multi-armed bandit approach to nonparametric H-infinity-norm estimation
2017 (English)In: 2-s2.0-85046136421, Institute of Electrical and Electronics Engineers (IEEE), 2017, p. 4632-4637Conference paper, Published paper (Refereed)
Abstract [en]

We study the problem of estimating the largest gain of an unknown linear and time-invariant filter, which is also known as the H-infinity norm of the system. By using ideas from the stochastic multi-armed bandit framework, we present a new algorithm that sequentially designs an input signal in order to estimate this quantity by means of input-output data. The algorithm is shown empirically to beat an asymptotically optimal method, known as Thompson Sampling, in the sense of its cumulative regret function. Finally, for a general class of algorithms, a lower bound on the performance of finding the H-infinity norm is derived.

Place, publisher, year, edition, pages
Institute of Electrical and Electronics Engineers (IEEE), 2017
Series
IEEE Conference on Decision and Control, ISSN 0743-1546
National Category
Electrical Engineering, Electronic Engineering, Information Engineering
Identifiers
urn:nbn:se:kth:diva-223861 (URN)10.1109/CDC.2017.8264343 (DOI)000424696904075 ()2-s2.0-85046136421 (Scopus ID)978-1-5090-2873-3 (ISBN)
Conference
IEEE 56th Annual Conference on Decision and Control (CDC), DEC 12-15, 2017, Melbourne, AUSTRALIA
Funder
Swedish Research Council, 2015-04393; 2016-06079
Note

QC 20180306

Available from: 2018-03-06 Created: 2018-03-06 Last updated: 2018-06-04Bibliographically approved
Mattila, R., Rojas, C. R., Krishnamurthy, V. & Wahlberg, B. (2017). Asymptotically Efficient Identification of Known-Sensor Hidden Markov Models. IEEE Signal Processing Letters, 24(12), 1813-1817
Open this publication in new window or tab >>Asymptotically Efficient Identification of Known-Sensor Hidden Markov Models
2017 (English)In: IEEE Signal Processing Letters, ISSN 1070-9908, E-ISSN 1558-2361, Vol. 24, no 12, p. 1813-1817Article in journal (Refereed) Published
Abstract [en]

We consider estimating the transition probability matrix of a finite-state finite-observation alphabet hidden Markov model with known observation probabilities. We propose a two-step algorithm: a method of moments estimator (formulated as a convex optimization problem) followed by a single iteration of a Newton-Raphson maximum-likelihood estimator. The two-fold contribution of this letter is, first, to theoretically show that the proposed estimator is consistent and asymptotically efficient, and second, to numerically show that the method is computationally less demanding than conventional methods-in particular for large datasets.

Place, publisher, year, edition, pages
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC, 2017
Keywords
Hidden Markov models (HMM), maximum-likelihood (ML), method of moments, system identification
National Category
Electrical Engineering, Electronic Engineering, Information Engineering
Identifiers
urn:nbn:se:kth:diva-217930 (URN)10.1109/LSP.2017.2759902 (DOI)000413962800006 ()2-s2.0-85031786373 (Scopus ID)
Note

QC 20171121

Available from: 2017-11-21 Created: 2017-11-21 Last updated: 2017-11-21Bibliographically approved
Eckhard, D., Bazanella, A. S., Rojas, C. R. & Hjalmarsson, H. (2017). Cost function shaping of the output error criterion. Automatica, 76, 53-60
Open this publication in new window or tab >>Cost function shaping of the output error criterion
2017 (English)In: Automatica, ISSN 0005-1098, E-ISSN 1873-2836, Vol. 76, p. 53-60Article in journal (Refereed) Published
Abstract [en]

Identification of an output error model using the prediction error method leads to an optimization problem built on input/output data collected from the system to be identified. It is often hard to find the global solution of this optimization problem because in most cases both the corresponding objective function and the search space are nonconvex. The difficulty in solving the optimization problem depends mainly on the experimental conditions, more specifically on the spectra of the input/output data collected from the system. It is therefore possible to improve the convergence of the algorithms by properly choosing the data prefilters; in this paper we show how to perform this choice. We present the application of the proposed approach to case studies where the standard algorithms tend to fail to converge to the global minimum.

Place, publisher, year, edition, pages
Elsevier, 2017
Keywords
Identification methods, Model fitting
National Category
Control Engineering
Identifiers
urn:nbn:se:kth:diva-202642 (URN)10.1016/j.automatica.2016.10.015 (DOI)000392788100007 ()2-s2.0-85001975561 (Scopus ID)
Note

QC 20170307

Available from: 2017-03-07 Created: 2017-03-07 Last updated: 2017-11-29Bibliographically approved
Eckhard, D., Bazanella, A. S., Rojas, C. R. & Hjalmarsson, H. (2017). Cost function shaping of the output error criterion. Automatica, 76, 53-60
Open this publication in new window or tab >>Cost function shaping of the output error criterion
2017 (English)In: Automatica, ISSN 0005-1098, E-ISSN 1873-2836, Vol. 76, p. 53-60Article in journal (Refereed) Published
Abstract [en]

Identification of an output error model using the prediction error method leads to an optimization problem built on input/output data collected from the system to be identified. It is often hard to find the global solution of this optimization problem because in most cases both the corresponding objective function and the search space are nonconvex. The difficulty in solving the optimization problem depends mainly on the experimental conditions, more specifically on the spectra of the input/output data collected from the system. It is therefore possible to improve the convergence of the algorithms by properly choosing the data prefilters; in this paper we show how to perform this choice. We present the application of the proposed approach to case studies where the standard algorithms tend to fail to converge to the global minimum.

Place, publisher, year, edition, pages
PERGAMON-ELSEVIER SCIENCE LTD, 2017
Keywords
Identification methods, Model fitting
National Category
Control Engineering
Identifiers
urn:nbn:se:kth:diva-202774 (URN)10.1016/j.automatica.2016.10.015 (DOI)000392788100007 ()
Note

QC 20170307

Available from: 2017-03-07 Created: 2017-03-07 Last updated: 2017-06-29Bibliographically approved
Rallo, G., Formentin, S., Rojas, C. R., Oomen, T. & Savaresi, S. M. (2017). Data-driven H-infinity-norm estimation via expert advice. In: 2017 IEEE 56TH ANNUAL CONFERENCE ON DECISION AND CONTROL (CDC): . Paper presented at IEEE 56th Annual Conference on Decision and Control (CDC), DEC 12-15, 2017, Melbourne, AUSTRALIA. IEEE
Open this publication in new window or tab >>Data-driven H-infinity-norm estimation via expert advice
Show others...
2017 (English)In: 2017 IEEE 56TH ANNUAL CONFERENCE ON DECISION AND CONTROL (CDC), IEEE , 2017Conference paper, Published paper (Refereed)
Abstract [en]

H-infinity-norm estimation is usually an important aspect of robust control design. The aim of this paper is to develop a data-driven estimation method exploiting iterative input design, without requiring parametric modeling. More specifically, the estimation problem is formulated as a sequential game, whose solution is derived within the prediction with expert advice framework. The proposed method is shown to be competitive with the state-of-the-art techniques.

Place, publisher, year, edition, pages
IEEE, 2017
Series
IEEE Conference on Decision and Control, ISSN 0743-1546
National Category
Electrical Engineering, Electronic Engineering, Information Engineering
Identifiers
urn:nbn:se:kth:diva-223867 (URN)000424696901091 ()978-1-5090-2873-3 (ISBN)
Conference
IEEE 56th Annual Conference on Decision and Control (CDC), DEC 12-15, 2017, Melbourne, AUSTRALIA
Note

QC 20180306

Available from: 2018-03-06 Created: 2018-03-06 Last updated: 2018-03-06Bibliographically approved
Mattila, R., Rojas, C. R., Krishnamurthy, V. & Wahlberg, B. (2017). Identification of Hidden Markov Models Using Spectral Learning with Likelihood Maximization. In: 2017 IEEE 56th Annual Conference on Decision and Control, CDC 2017: . Paper presented at IEEE 56th Annual Conference on Decision and Control (CDC), DEC 12-15, 2017, Melbourne, Australia (pp. 5859-5864). Institute of Electrical and Electronics Engineers (IEEE)
Open this publication in new window or tab >>Identification of Hidden Markov Models Using Spectral Learning with Likelihood Maximization
2017 (English)In: 2017 IEEE 56th Annual Conference on Decision and Control, CDC 2017, Institute of Electrical and Electronics Engineers (IEEE), 2017, p. 5859-5864Conference paper, Published paper (Refereed)
Abstract [en]

In this paper, we consider identifying a hidden Markov model (HMM) with the purpose of computing estimates of joint and conditional (posterior) probabilities over observation sequences. The classical maximum likelihood estimation algorithm (via the Baum-Welch/expectation-maximization algorithm), has recently been challenged by methods of moments. Such methods employ low-order moments to provide parameter estimates and have several benefits, including consistency and low computational cost. This paper aims to reduce the gap in statistical efficiency that results from restricting to only low-order moments in the training data. In particular, we propose a two-step procedure that combines spectral learning with a single Newton-like iteration for maximum likelihood estimation. We demonstrate an improved statistical performance using the proposed algorithm in numerical simulations.

Place, publisher, year, edition, pages
Institute of Electrical and Electronics Engineers (IEEE), 2017
Series
IEEE Conference on Decision and Control, ISSN 0743-1546
National Category
Electrical Engineering, Electronic Engineering, Information Engineering
Identifiers
urn:nbn:se:kth:diva-223860 (URN)10.1109/CDC.2017.8264545 (DOI)000424696905103 ()2-s2.0-85046135167 (Scopus ID)978-1-5090-2873-3 (ISBN)
Conference
IEEE 56th Annual Conference on Decision and Control (CDC), DEC 12-15, 2017, Melbourne, Australia
Funder
Swedish Research Council, 2016-06079
Note

QC 20180306

Available from: 2018-03-06 Created: 2018-03-06 Last updated: 2018-06-01Bibliographically approved
Organisations
Identifiers
ORCID iD: ORCID iD iconorcid.org/0000-0003-0355-2663

Search in DiVA

Show all publications