Search

In the last decade, kernel-based regularization methods (KRMs) have been widely used for stable impulse response estimation in system identification. Its favorable performance over classic maximum likelihood/prediction error methods (ML/PEM) has been verified by extensive simulations. Recently, we noticed a surprising observation: for some data sets and kernels, no matter how the hyper-parameters are tuned, the regularized least square estimate cannot have higher model fit than the least square (LS) estimate, which implies that for such cases, the regularization cannot improve the LS estimate. Therefore, this paper focuses on how to understand this observation. To this purpose, we first introduce the squared error (SE) criterion, and the corresponding oracle hyper-parameter estimator in the sense of minimizing the SE criterion. Then we find the necessary and sufficient conditions under which the regularization cannot improve the LS estimate, and we show that the probability that this happens is greater than zero. The theoretical findings are demonstrated through numerical simulations, and simultaneously the anomalous simulation outcome wherein the probability is nearly zero is elucidated, and due to the ill-conditioned nature of either the kernel matrix, the Gram matrix, or both. (c) 2023 Elsevier Ltd. All rights reserved., Funding Agencies|National Key R&D Program of China [2018YFA0703800]; National Natural Science Foundation of China [62273287]; Shenzhen Science and Technology Innovation Council [JCYJ20220530143418040]; Thousand Youth Talents Plan - central government of China

We consider signal source localization from range-difference measurements. First, we give some readily-checked conditions on measurement noises and sensor deployment to guarantee the asymptotic identifiability of the model and show the consistency and asymptotic normality of the maximum likelihood (ML) estimator. Then, we devise an estimator that owns the same asymptotic property as the ML one. Specifically, we prove that the negative log-likelihood function converges to a function, which has a unique minimum and positive definite Hessian at the true source’s position. Hence, it is promising to execute local iterations, e.g., the Gauss-Newton (GN) algorithm, following a consistent estimate. The main issue involved is obtaining a preliminary consistent estimate. To this aim, we construct a linear least-squares problem via algebraic operation and constraint relaxation and obtain a closed-form solution. We then focus on deriving and eliminating the bias of the linear least-squares estimator, which yields an asymptotically unbiased and further consistent estimate. Noting that the bias is a function of the noise variance, we further devise a consistent noise variance estimator that involves a 3-order polynomial rooting. Based on the preliminary consistent location estimate, a one-step GN iteration suffices to achieve the same asymptotic property as the ML estimator. Simulation results demonstrate the superiority of our proposed algorithm in the large sample case. IEEE

Given 2D point correspondences between an image pair, inferring the camera motion is a fundamental issue in the computer vision community. The existing works generally set out from the epipolar constraint and estimate the essential matrix, which is not optimal in the maximum likelihood (ML) sense. In this paper, we dive into the original measurement model with respect to the rotation matrix and normalized translation vector and formulate the ML problem. We then propose a two-step algorithm to solve it: In the first step, we estimate the variance of measurement noises and devise a consistent estimator based on bias elimination; In the second step, we execute a one-step Gauss-Newton iteration on manifold to refine the consistent estimate. We prove that the proposed estimate owns the same asymptotic statistical properties as the ML estimate: The first is consistency, i.e., the estimate converges to the ground truth as the point number increases; The second is asymptotic efficiency, i.e., the mean squared error of the estimate converges to the theoretical lower bound -- Cramer-Rao bound. In addition, we show that our algorithm has linear time complexity. These appealing characteristics endow our estimator with a great advantage in the case of dense point correspondences. Experiments on both synthetic data and real images demonstrate that when the point number reaches the order of hundreds, our estimator outperforms the state-of-the-art ones in terms of estimation accuracy and CPU time.

For data-driven iterative learning control (ILC) methods, both the model estimation and controller design problems are converted to parameter estimation problems for some chosen model structures. It is well-known that if the model order is not chosen carefully, models with either large variance or large bias would be resulted, which is one of the obstacles to further improve the modeling and tracking performances of data-driven ILC in practice. An emerging trend in the system identification community to deal with this issue is using regularization instead of the statistical tests, e.g., AIC, BIC, and one of the representatives is the so-called kernel-based regularization method (KRM). In this paper, we integrate KRM into data-driven ILC to handle a class of repetitive linear time-varying systems, and moreover, we show that the proposed method has ultimately bounded tracking error in the iteration domain. The numerical simulation results show that in contrast with the least squares method and some existing data-driven ILC methods, the proposed one can give faster convergence speed, better accuracy and robustness in terms of the tracking performance., Comment: 17 pages

Regularized techniques, also named as kernel-based techniques, are the major advances in system identification in the last decade. Although many promising results have been achieved, their theoretical analysis is far from complete and there are still many key problems to be solved. One of them is the asymptotic theory, which is about convergence properties of the model estimators as the sample size goes to infinity. The existing related results for regularized system identification are about the almost sure convergence of various hyperparameter estimators. A common problem of those results is that they do not contain information on the factors that affect the convergence properties of those hyperparameter estimators, e.g., the regression matrix. In this article, we tackle problems of this kind for the regularized finite impulse response model estimation with the empirical Bayes (EB) hyperparameter estimator and filtered white noise input. In order to expose and find those factors, we study the convergence in distribution of the EB hyperparameter estimator, and the asymptotic distribution of its corresponding model estimator. For illustration, we run Monte Carlo simulations to show the efficacy of our obtained theoretical results., Funding Agencies|National Natural Science Foundation of China

Asymptotic theory for the regularized system identification has received increasing interests in recent years. In this paper, for the finite impulse response (FIR) model and filtered white noise inputs, we show the convergence in distribution of the Steins unbiased risk estimator (SURE) based hyper-parameter estimator and find factors that influence its convergence properties. In particular, we consider the ridge regression case to obtain closed-form expressions of the limit of the regression matrix and the variance of the limiting distribution of the SURE based hyper-parameter estimator, and then demonstrate their relation numerically., Funding Agencies|Thousand Youth Talents Plan funded by the central government of China - NSFC [61773329]; Shenzhen Science and Technology Innovation Council [Ji-20170189, JCY20170411102101881]; Robotic Discipline Development Fund from Shenzhen Government [2016-1418]; CUHKSZ [PF. 01.000249, 2014.0003.23]; Swedish Research Council [2019-04956]; Vinnovas center LINKSIC

Input design is an important problem for system identification and has been well studied for the classical system identification, i.e., the maximum likelihood/prediction error method. For the emerging regularized system identification, the study on input design has just started, and it is often formulated as a non-convex optimization problem that minimizes a scalar measure of the Bayesian mean squared error matrix subject to certain constraints, and the state-of-art method is the so-called quadratic mapping and inverse embedding (QMIE) method, where a time domain inverse embedding (TDIE) is proposed to find the inverse of the quadratic mapping. In this paper, we report some new results on the embeddings/inverse embeddings of the QMIE method. Firstly, we present a general result on the frequency domain inverse embedding (FDIE) that is to find the inverse of the quadratic mapping described by the discrete-time Fourier transform. Then we show the relation between the TDIE and the FDIE from a graph signal processing perspective. Finally, motivated by this perspective, we further propose a graph induced embedding and its inverse, which include the previously introduced embeddings as special cases. This deepens the understanding of input design from a new viewpoint beyond the real domain and the frequency domain viewpoints.

Regularized system identification is the major advance in system identification in the last decade. Although many promising results have been achieved, it is far from complete and there are still many key problems to be solved. One of them is the asymptotic theory, which is about convergence properties of the model estimators as the sample size goes to infinity. The existing related results for regularized system identification are about the almost sure convergence of various hyper-parameter estimators. A common problem of those results is that they do not contain information on the factors that affect the convergence properties of those hyper-parameter estimators, e.g., the regression matrix. In this paper, we tackle problems of this kind for the regularized finite impulse response model estimation with the empirical Bayes (EB) hyper-parameter estimator and filtered white noise input. In order to expose and find those factors, we study the convergence in distribution of the EB hyper-parameter estimator, and the asymptotic distribution of its corresponding model estimator. For illustration, we run Monte Carlo simulations to show the efficacy of our obtained theoretical results.

Spatial-temporal Gaussian process regression is a popular method for spatial-temporal data modeling. Its state-of-art implementation is based on the state-space model realization of the spatial-temporal Gaussian process and its corresponding Kalman filter and smoother, and has computational complexity $\mathcal{O}(NM^3)$, where $N$ and $M$ are the number of time instants and spatial input locations, respectively, and thus can only be applied to data with large $N$ but relatively small $M$. In this paper, our primary goal is to show that by exploring the Kronecker structure of the state-space model realization of the spatial-temporal Gaussian process, it is possible to further reduce the computational complexity to $\mathcal{O}(M^3+NM^2)$ and thus the proposed implementation can be applied to data with large $N$ and moderately large $M$. The proposed implementation is illustrated over applications in weather data prediction and spatially-distributed system identification. Our secondary goal is to design a kernel for both the Colorado precipitation data and the GHCN temperature data, such that while having more efficient implementation, better prediction performance can also be achieved than the state-of-art result.

State estimation is an essential part of autonomous systems. Integrating the Ultra-Wideband(UWB) technique has been shown to correct the long-term estimation drift and bypass the complexity of loop closure detection. However, few works on robotics adopt UWB as a stand-alone state estimation solution. The primary purpose of this work is to investigate planar pose estimation using only UWB range measurements and study the estimator's statistical efficiency. We prove the excellent property of a two-step scheme, which says that we can refine a consistent estimator to be asymptotically efficient by one step of Gauss-Newton iteration. Grounded on this result, we design the GN-ULS estimator and evaluate it through simulations and collected datasets. GN-ULS attains millimeter and sub-degree level accuracy on our static datasets and attains centimeter and degree level accuracy on our dynamic datasets, presenting the possibility of using only UWB for real-time state estimation., Comment: Update the content and improve consistency with the ICRA version

The Perspective-n-Point (PnP) problem has been widely studied in both computer vision and photogrammetry societies. With the development of feature extraction techniques, a large number of feature points might be available in a single shot. It is promising to devise a consistent estimator, i.e., the estimate can converge to the true camera pose as the number of points increases. To this end, we propose a consistent PnP solver, named \emph{CPnP}, with bias elimination. Specifically, linear equations are constructed from the original projection model via measurement model modification and variable elimination, based on which a closed-form least-squares solution is obtained. We then analyze and subtract the asymptotic bias of this solution, resulting in a consistent estimate. Additionally, Gauss-Newton (GN) iterations are executed to refine the consistent solution. Our proposed estimator is efficient in terms of computations -- it has $O(n)$ computational complexity. Experimental tests on both synthetic data and real images show that our proposed estimator is superior to some well-known ones for images with dense visual features, in terms of estimation precision and computing time.

Asymptotic theory is one of the core subjects in system identification theory and often used to assess properties of model estimators. In this paper, we focus on the asymptotic theory for the kernel-based regularized system identification and study the convergence in distribution of the generalized cross validation (GCV) based hyper-parameter estimator. It is shown that the difference between the GCV based hyper-parameter estimator and the optimal hyper-parameter estimator that minimizes the mean square error scaled by 1/root N converges in distribution to a zero mean Gaussian distribution, where N is the sample size and an expression of covariance matrix is obtained. In particular, for the ridge regression case, a closed-form expression of the variance is obtained and shows the influence of the limit of the regression matrix on the asymptotic distribution. For illustration, Monte Carlo numerical simulations are run to test our theoretical results., Funding Agencies|Thousand Youth Talents Plan funded by the central government of China - NSFC [61773329]; Shenzhen Science and Technology Innovation Council [Ji-20170189, JCY20170411102101881]; Robotic Discipline Development Fund [20161418]; Shenzhen Government [2014.0003.23]; CUHKSZ; Swedish Research Council [2019-04956]

In this paper, we give a tutorial on asymptotic properties of the Least Square (LS) and Regularized Least Squares (RLS) estimators for the finite impulse response model with filtered white noise inputs. We provide three perspectives: the almost sure convergence, the convergence in distribution and the boundedness in probability. On one hand, these properties deepen our understanding of the LS and RLS estimators. On the other hand, we can use them as tools to investigate asymptotic properties of other estimators, such as various hyper-parameter estimators.

This paper investigates two issues on identification of switched linear systems: persistence of excitation and numerical algorithms. The main contribution is a much weaker condition on the regressor to be persistently exciting that guarantees the uniqueness of the parameter sets and also provides new insights in understanding the relation among different subsystems. It is found that for uniquely determining the parameters of switched linear systems, the minimum number of samples needed derived from our condition is much smaller than that reported in the literature. The secondary contribution of the paper concerns the numerical algorithm. Though the algorithm is not new, we show that our surrogate problem, relaxed from an integer optimization to a continuous minimization, has exactly the same solution as the original integer optimization, which is effectively solved by a block-coordinate descent algorithm. Moreover, an algorithm for handling unknown number of subsystems is proposed. Several numerical examples are illustrated to support theoretical analysis.

The selections for the model orders and the number of controller parameters have not been discussed for many data-driven iterative learning control (ILC) methods. If they are not chosen carefully, the estimated model and designed controller will lead to either large variance or large bias. In this paper we try to use the kernel-based regularization method (KRM) to handle the model estimation problem and the controller design problem for unknown repetitive linear time-varying systems. In particular, we have used the diagonal correlated kernel and the marginal likelihood maximization method for the two problems. Numerical simulation results show that smaller mean square errors for each time instant are obtained by using the proposed ILC method in comparison with an existing data-driven ILC approach. Copyright (C) 2021 The Authors., Funding Agencies|Thousand Youth Talents Plan funded by the central government of China - NSFCNational Natural Science Foundation of China (NSFC) [61773329]; Shenzhen Science and Technology Innovation Council [Ji20170189]; CUHKSZ; Strategic Priority Research Program of Chinese Academy of SciencesChinese Academy of Sciences [XDA27000000]; Swedish Research CouncilSwedish Research CouncilEuropean Commission [2019-04956]; Vinnovas center LINKSICVinnova; [2014.0003.23]

System identification is a mature research area with well established paradigms, mostly based on classical statistical methods. Recently, there has been considerable interest in so called kernel-based regularisation methods applied to system identification problem. The recent literature on this is extensive and at times difficult to digest. The purpose of this contribution is to provide an accessible account of the main ideas and results of kernel-based regularisation methods for system identification. The focus is to assess the impact of these new techniques on the field and traditional paradigms.

System identification is a mature research area with well established paradigms, mostly based on classical statistical methods. Recently, there has been considerable interest in so called kernel-based regularisation methods applied to system identification problem. The recent literature on this is extensive and at times difficult to digest. The purpose of this contribution is to provide an accessible account of the main ideas and results of kernel-based regularisation methods for system identification. The focus is to assess the impact of these new techniques on the field and traditional paradigms.

System identification is a mature research area with well established paradigms, mostly based on classical statistical methods. Recently, there has been considerable interest in so called kernel-based regularisation methods applied to system identification problem. The recent literature on this is extensive and at times difficult to digest. The purpose of this contribution is to provide an accessible account of the main ideas and results of kernel-based regularisation methods for system identification. The focus is to assess the impact of these new techniques on the field and traditional paradigms.

System identification is a mature research area with well established paradigms, mostly based on classical statistical methods. Recently, there has been considerable interest in so called kernel-based regularisation methods applied to system identification problem. The recent literature on this is extensive and at times difficult to digest. The purpose of this contribution is to provide an accessible account of the main ideas and results of kernel-based regularisation methods for system identification. The focus is to assess the impact of these new techniques on the field and traditional paradigms.

In this paper, we focus on the influences of the condition number of the regression matrix upon the comparison between two hyper-parameter estimation methods: the empirical Bayes (EB) and the Stein's unbiased estimator with respect to the mean square error (MSE) related to output prediction (SUREy). We firstly show that the greatest power of the condition number of the regression matrix of SUREy cost function convergence rate upper bound is always one larger than that of EB cost function convergence rate upper bound. Meanwhile, EB and SUREy hyper-parameter estimators are both proved to be asymptotically normally distributed under suitable conditions. In addition, one ridge regression case is further investigated to show that when the condition number of the regression matrix goes to infinity, the asymptotic variance of SUREy estimator tends to be larger than that of EB estimator.

In this paper, we study the influence of ill-conditioned regression matrix on two hyper-parameter estimation methods for the kernel-based regularization method: the empirical Bayes (EB) and the Steins unbiased risk estimator (SURE). First, we consider the convergence rate of the cost functions of EB and SURE, and we find that they have the same convergence rate but the influence of the ill-conditioned regression matrix on the scale factor are different: for upper bounds, the scale factor for SURE contains one more factor cond(Phi(T)Phi) than that of EB, where Phi is the regression matrix and cond(.) denotes the condition number of a matrix. This finding indicates that when Phi is ill-conditioned, i.e., cond(Phi(T)Phi) is large, the cost function of SURE converges slower than that of EB. Then we consider the convergence rate of the optimal hyper-parameters of EB and SURE, and we find that they are both asymptotically normally distributed and have the same convergence rate, but the influence of the ill-conditioned regression matrix on the scale factor are different. In particular, for the ridge regression case, we show that the optimal hyper-parameter of SURE converges slower than that of EB with a factor of 1/n(2), as cond(Phi(T)Phi) goes to infinity, where n is the FIR model order., Funding Agencies|Thousand Youth Talents Plan funded by the central government of China - NSFCNational Natural Science Foundation of China (NSFC) [61773329]; Shenzhen Science and Technology Innovation Council [Ji-20170189 (J-CY20170411102101881)]; Robotic Discipline Development Fund from Shenzhen Government [20161418]; CUHKSZ [2014.0003.23]; [PF. 01.000249]

System identification is a mature research area with well established paradigms, mostly based on classical statistical methods. Recently, there has been considerable interest in so called kernel-based regularisation methods applied to system identification problem. The recent literature on this is extensive and at times difficult to digest. The purpose of this contribution is to provide an accessible account of the main ideas and results of kernel-based regularisation methods for system identification. The focus is to assess the impact of these new techniques on the field and traditional paradigms.

In this paper, we study the recursion of measurement outcomes for open quantum networks under sequential measurements. Open quantum networks are networked quantum subsystems (e.g., qubits) with the state evolutions described by a continuous Lindblad master equation. When measurements are performed sequentially along such continuous dynamics, the quantum network states undergo random jumps and the corresponding measurement outcomes can be described by a vector of probabilistic Boolean variables. The induced recursion of the Boolean vectors forms a probabilistic Boolean network. First of all, we show that the state transition of the induced Boolean networks can be explicitly represented through realification of the master equation. Next, when the open quantum dynamics is relaxing in the sense that it possesses a unique equilibrium as a global attractor, structural properties including absorbing states, reducibility, and periodicity for the induced Boolean network are direct consequences of the relaxing property. Particularly, we show that generically, relaxing quantum dynamics leads to irreducible and aperiodic chains for the measurement outcomes. Finally, we show that for quantum consensus networks as a type of non-relaxing open quantum network dynamics, the communication classes of the measurement-induced Boolean networks are encoded in the quantum Laplacian of the underlying interaction graph., Comment: 21 pages, 3 figures

In this paper, we study dynamical quantum networks which evolve according to Schr\"odinger equations but subject to sequential local or global quantum measurements. A network of qubits forms a composite quantum system whose state undergoes unitary evolution in between periodic measurements, leading to hybrid quantum dynamics with random jumps at discrete time instances along a continuous orbit. The measurements either act on the entire network of qubits, or only a subset of qubits. First of all, we reveal that this type of hybrid quantum dynamics induces probabilistic Boolean recursions representing the measurement outcomes. With global measurements, it is shown that such resulting Boolean recursions define Markov chains whose state-transitions are fully determined by the network Hamiltonian and the measurement observables. Particularly, we establish an explicit and algebraic representation of the underlying recursive random mapping driving such induced Markov chains. Next, with local measurements, the resulting probabilistic Boolean dynamics is shown to be no longer Markovian. The state transition probability at any given time becomes dependent on the entire history of the sample path, for which we establish a recursive way of computing such non-Markovian probability transitions. Finally, we adopt the classical bilinear control model for the continuous Schr\"odinger evolution, and show how the measurements affect the controllability of the quantum networks., Comment: 26 pages, 7 figures

The kernel-based regularization method has two core issues: kernel design and hyperparameter estimation. In this paper, we focus on the second issue and study the properties of several hyperparameter estimators including the empirical Bayes (EB) estimator, two Steins unbiased risk estimators (SURE) (one related to impulse response reconstruction and the other related to output prediction) and their corresponding Oracle counterparts, with an emphasis on the asymptotic properties of these hyperparameter estimators. To this goal, we first derive and then rewrite the first order optimality conditions of these hyperparameter estimators, leading to several insights on these hyperparameter estimators. Then we show that as the number of data goes to infinity, the two SUREs converge to the best hyperparameter minimizing the corresponding mean square error, respectively, while the more widely used EB estimator converges to another best hyperparameter minimizing the expectation of the EB estimation criterion. This indicates that the two SUREs are asymptotically optimal in the corresponding MSE senses but the EB estimator is not. Surprisingly, the convergence rate of two SUREs is slower than that of the EB estimator, and moreover, unlike the two SUREs, the EB estimator is independent of the convergence rate of Phi(T)Phi/N to its limit, where Phi is the regression matrix and N is the number of data. A Monte Carlo simulation is provided to demonstrate the theoretical results. (C) 2018 Elsevier Ltd. All rights reserved., Funding Agencies|National Natural Science Foundation of China [61773329, 61603379]; central government of China; Shenzhen Science and Technology Innovation Council [Ji-20170189, Ji-20160207]; Chinese University of Hong Kong, Shenzhen [PF. 01.000249, 2014.0003.23]; Swedish Research Council [2014-5894]; National Key Basic Research Program of China (973 Program) [2014CB845301]; Presidential Fund of the Academy of Mathematics and Systems Science, CAS [2015-hwyxqnrc-mbq]

The kernel-based regularization method has two core issues: kernel design and hyperparameter estimation. In this paper, we focus on the second issue and study the properties of several hyperparameter estimators including the empirical Bayes (EB) estimator, two Steins unbiased risk estimators (SURE) (one related to impulse response reconstruction and the other related to output prediction) and their corresponding Oracle counterparts, with an emphasis on the asymptotic properties of these hyperparameter estimators. To this goal, we first derive and then rewrite the first order optimality conditions of these hyperparameter estimators, leading to several insights on these hyperparameter estimators. Then we show that as the number of data goes to infinity, the two SUREs converge to the best hyperparameter minimizing the corresponding mean square error, respectively, while the more widely used EB estimator converges to another best hyperparameter minimizing the expectation of the EB estimation criterion. This indicates that the two SUREs are asymptotically optimal in the corresponding MSE senses but the EB estimator is not. Surprisingly, the convergence rate of two SUREs is slower than that of the EB estimator, and moreover, unlike the two SUREs, the EB estimator is independent of the convergence rate of Phi(T)Phi/N to its limit, where Phi is the regression matrix and N is the number of data. A Monte Carlo simulation is provided to demonstrate the theoretical results. (C) 2018 Elsevier Ltd. All rights reserved., Funding Agencies|National Natural Science Foundation of China [61773329, 61603379]; central government of China; Shenzhen Science and Technology Innovation Council [Ji-20170189, Ji-20160207]; Chinese University of Hong Kong, Shenzhen [PF. 01.000249, 2014.0003.23]; Swedish Research Council [2014-5894]; National Key Basic Research Program of China (973 Program) [2014CB845301]; Presidential Fund of the Academy of Mathematics and Systems Science, CAS [2015-hwyxqnrc-mbq]

The kernel-based regularization method has two core issues: kernel design and hyperparameter estimation. In this paper, we focus on the second issue and study the properties of several hyperparameter estimators including the empirical Bayes (EB) estimator, two Steins unbiased risk estimators (SURE) (one related to impulse response reconstruction and the other related to output prediction) and their corresponding Oracle counterparts, with an emphasis on the asymptotic properties of these hyperparameter estimators. To this goal, we first derive and then rewrite the first order optimality conditions of these hyperparameter estimators, leading to several insights on these hyperparameter estimators. Then we show that as the number of data goes to infinity, the two SUREs converge to the best hyperparameter minimizing the corresponding mean square error, respectively, while the more widely used EB estimator converges to another best hyperparameter minimizing the expectation of the EB estimation criterion. This indicates that the two SUREs are asymptotically optimal in the corresponding MSE senses but the EB estimator is not. Surprisingly, the convergence rate of two SUREs is slower than that of the EB estimator, and moreover, unlike the two SUREs, the EB estimator is independent of the convergence rate of Phi(T)Phi/N to its limit, where Phi is the regression matrix and N is the number of data. A Monte Carlo simulation is provided to demonstrate the theoretical results. (C) 2018 Elsevier Ltd. All rights reserved., Funding Agencies|National Natural Science Foundation of China [61773329, 61603379]; central government of China; Shenzhen Science and Technology Innovation Council [Ji-20170189, Ji-20160207]; Chinese University of Hong Kong, Shenzhen [PF. 01.000249, 2014.0003.23]; Swedish Research Council [2014-5894]; National Key Basic Research Program of China (973 Program) [2014CB845301]; Presidential Fund of the Academy of Mathematics and Systems Science, CAS [2015-hwyxqnrc-mbq]

The kernel-based regularization method has two core issues: kernel design and hyperparameter estimation. In this paper, we focus on the second issue and study the properties of several hyperparameter estimators including the empirical Bayes (EB) estimator, two Steins unbiased risk estimators (SURE) (one related to impulse response reconstruction and the other related to output prediction) and their corresponding Oracle counterparts, with an emphasis on the asymptotic properties of these hyperparameter estimators. To this goal, we first derive and then rewrite the first order optimality conditions of these hyperparameter estimators, leading to several insights on these hyperparameter estimators. Then we show that as the number of data goes to infinity, the two SUREs converge to the best hyperparameter minimizing the corresponding mean square error, respectively, while the more widely used EB estimator converges to another best hyperparameter minimizing the expectation of the EB estimation criterion. This indicates that the two SUREs are asymptotically optimal in the corresponding MSE senses but the EB estimator is not. Surprisingly, the convergence rate of two SUREs is slower than that of the EB estimator, and moreover, unlike the two SUREs, the EB estimator is independent of the convergence rate of Phi(T)Phi/N to its limit, where Phi is the regression matrix and N is the number of data. A Monte Carlo simulation is provided to demonstrate the theoretical results. (C) 2018 Elsevier Ltd. All rights reserved., Funding Agencies|National Natural Science Foundation of China [61773329, 61603379]; central government of China; Shenzhen Science and Technology Innovation Council [Ji-20170189, Ji-20160207]; Chinese University of Hong Kong, Shenzhen [PF. 01.000249, 2014.0003.23]; Swedish Research Council [2014-5894]; National Key Basic Research Program of China (973 Program) [2014CB845301]; Presidential Fund of the Academy of Mathematics and Systems Science, CAS [2015-hwyxqnrc-mbq]

The kernel-based regularization method has two core issues: kernel design and hyperparameter estimation. In this paper, we focus on the second issue and study the properties of several hyperparameter estimators including the empirical Bayes (EB) estimator, two Steins unbiased risk estimators (SURE) (one related to impulse response reconstruction and the other related to output prediction) and their corresponding Oracle counterparts, with an emphasis on the asymptotic properties of these hyperparameter estimators. To this goal, we first derive and then rewrite the first order optimality conditions of these hyperparameter estimators, leading to several insights on these hyperparameter estimators. Then we show that as the number of data goes to infinity, the two SUREs converge to the best hyperparameter minimizing the corresponding mean square error, respectively, while the more widely used EB estimator converges to another best hyperparameter minimizing the expectation of the EB estimation criterion. This indicates that the two SUREs are asymptotically optimal in the corresponding MSE senses but the EB estimator is not. Surprisingly, the convergence rate of two SUREs is slower than that of the EB estimator, and moreover, unlike the two SUREs, the EB estimator is independent of the convergence rate of Phi(T)Phi/N to its limit, where Phi is the regression matrix and N is the number of data. A Monte Carlo simulation is provided to demonstrate the theoretical results. (C) 2018 Elsevier Ltd. All rights reserved., Funding Agencies|National Natural Science Foundation of China [61773329, 61603379]; central government of China; Shenzhen Science and Technology Innovation Council [Ji-20170189, Ji-20160207]; Chinese University of Hong Kong, Shenzhen [PF. 01.000249, 2014.0003.23]; Swedish Research Council [2014-5894]; National Key Basic Research Program of China (973 Program) [2014CB845301]; Presidential Fund of the Academy of Mathematics and Systems Science, CAS [2015-hwyxqnrc-mbq]

This paper studies the asymptotic properties of the hyperparameter estimators including the leave-k-out cross validation (LKOCV) and r-fold cross validation (RFCV), and discloses their relation with the Steins unbiased risk estimators (SURE) as well as the mean squared error (MSE). It is shown that as the number of data goes to infinity, the LKOCV shares the same asymptotic best hyperparameter minimizing the MSE estimator as the SURE does if the input is bounded and the ratio between the training data and the whole data tends to zero. We illustrate the efficacy of the theoretical result by Monte Carlo simulations., Funding Agencies|National Natural Science Foundation of China [61603379, 61773329]; National Key Basic Research Program of China (973 Program) [2014CB845301]; President Fund of Academy of Mathematics and Systems Science, CAS [2015-hwyxqnrc-mbq]; Shenzhen Science and Technology Innovation Council [Ji-20170189, Ji-20160207]; Chinese University of Hong Kong, Shenzhen [2014.0003.23]; Swedish Research Council [2014-5894]; Thousand Youth Talents Plan - central government of China; [PF. 01.000249]

Input design is an important issue for classical system identification methods but has not been investigated for the kernel-based regularization method (KRM) until very recently. In this paper, we consider the input design problem of KRMs for LTI system identification. Different from the recent result, we adopt a Bayesian perspective and in particular make use of scalar measures (e.g., the A-optimality, D-optimality, and E-optimality) of the Bayesian mean square error matrix as the design criteria subject to power-constraint on the input. Instead of solving the optimization problem directly, we propose a two-step procedure. In the first step, by making suitable assumptions on the unknown input, we construct a quadratic map (transformation) of the input such that the transformed input design problems are convex, and the global minima of the transformed input design problem can thus be found efficiently by applying well-developed convex optimization software packages. In the second step, we derive the characterization of the optimal input based on the global minima found in the first step by solving the inverse image of the quadratic map. In addition, we derive analytic results for some special types of kernels, which provide insights on the input design and also its dependence on the kernel structure. (C) 2018 Elsevier Ltd. All rights reserved., Funding Agencies|National Natural Science Foundation of China [61773329, 61603379]; Thousand Youth Talents Plan - central government of China; Shenzhen Projects - Shenzhen Science and Technology Innovation Council, China [Ji-20170189, Ji-20160207]; Chinese University of Hong Kong, Shenzhen, China [PF. 01.000249, 2014.0003.23]; Swedish Research Council [20145894]; National Key Basic Research Program of China (973 Program) [2014CB845301]; President Fund of Academy of Mathematics and Systems Science, CAS, China [2015-hwyxqnrc-mbq]

In this work, we study the multiple kernel based regularized system identification with the hyper-parameter estimated by using the Steins unbiased risk estimators (SURE). To approach the problem, a QR factorization is first employed to compute SUREs objective function and its gradient in an efficient and accurate way. Then we propose an algorithm to solve the SURE problem, which contains two parts: the outer optimization part and the inner optimization part. For the outer optimization part, the coordinate descent algorithm is used and for the inner optimization part, the projection gradient algorithm is used. Finally, the efficacy of the proposed algorithm is demonstrated by numerical simulations. (C) 2018, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved., Funding Agencies|Thousand Youth Talents Plan - central government of China; NSFC [61773329, 61603379]; Shenzhen Science and Technology Innovation Council [Ji-20170189, Ji-20160207]; Chinese University of Hong Kong, Shenzhen [2014.0003.23]; Swedish Research Council [621-2014-5894]; National Key Basic Research Program of China (973 Program) [2014CB845301]; Fund of Academy of Mathematics and Systems Science, CAS [2015-hwyxqnrc-mbq]; [PF. 01.000249]

In this paper, we study the asymptotic properties of the generalized cross validation (GCV) hyperparameter estimator and establish its connection with the Steins unbiased risk estimators (SURE) as well as the mean squared error (MSE). It is shown that as the number of data goes to infinity, the GCV has the same asymptotic property as the SURE does and both of them converge to the best hyperparameter in the MSE sense. We illustrate the efficacy of the result by Monte Carlo simulations. (C) 2018, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved., Funding Agencies|National Natural Science Foundation of China [61603379, 61773329]; National Key Basic Research Program of China (973 Program) [2014CB845301]; President Fund of Academy of Mathematics and Systems Science, CAS [2015-hwyxqnrc-mbq]; Thousand Youth Talents Plan - central government of China; Shenzhen Research Projects - Shenzhen Science and Technology Innovation Council [Ji-20170189, Ji-20160207]; Presidential Fund - Chinese University of Hong Kong, Shenzhen [PF.01.000249]; Swedish Research Council [2014-5894]

Regularization methods with regularization matrix in quadratic form have received increasing attention. For those methods, the design and tuning of the regularization matrix are two key issues that are closely related. For systems with complicated dynamics, it would be preferable that the designed regularization matrix can bring the hyper-parameter estimation problem certain structure such that a locally optimal solution can be found efficiently. An example of this idea is to use the so-called multiple kernel Chen et al. (2014) for kernel-based regularization methods. In this paper, we propose to use the multiple regularization matrix for the filter-based regularization. Interestingly, the marginal likelihood maximization with the multiple regularization matrix is also a difference of convex programming problem, and a locally optimal solution could be found with sequential convex optimization techniques. (C) 2018, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved., Funding Agencies|Thousand Youth Talents Plan of China; NSFC [61773329, 61603379]; Shenzhen Science and Technology Innovation Council [Ji-20170189, Ji-20160207]; Chinese University of Hong Kong, Shenzhen [2014.0003.23]; Swedish Research Council [2014-5894]; National Key Basic Research Program of China (973 Program) [2014CB845301]; AMSS, CAS [2015-hwyxqnrc-mbq]; [PF.01.000249]

In this work, we study the multiple kernel based regularized system identification with the hyper-parameter estimated by using the Stein's unbiased risk estimators (SURE). To approach the problem, a QR factorization is first employed to compute SURE's objective function and its gradient in an efficient and accurate way. Then we propose an algorithm to solve the SURE problem, which contains two parts: the outer optimization part and the inner optimization part. For the outer optimization part, the coordinate descent algorithm is used and for the inner optimization part, the projection gradient algorithm is used. Finally, the efficacy of the proposed algorithm is demonstrated by numerical simulations. (C) 2018, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.

Regularization methods with regularization matrix in quadratic form have received increasing attention. For those methods, the design and tuning of the regularization matrix are two key issues that are closely related. For systems with complicated dynamics, it would be preferable that the designed regularization matrix can bring the hyper-parameter estimation problem certain structure such that a locally optimal solution can be found efficiently. An example of this idea is to use the so-called multiple kernel Chen et al. (2014) for kernel-based regularization methods. In this paper, we propose to use the multiple regularization matrix for the filter-based regularization. Interestingly, the marginal likelihood maximization with the multiple regularization matrix is also a difference of convex programming problem, and a locally optimal solution could be found with sequential convex optimization techniques.

Regularization of simple linear regression models for system identification is a recent much-studied problem. Several parameterizations ("kernels") of the regularization matrix have been suggested together with different ways of estimating ("tuning") its parameters. This contribution defines an asymptotic view on the problem of tuning and selection of kernels. It is shown that the SURE approach to parameter tuning provides an asymptotically consistent estimate of the optimal (in a MSE sense) hyperparameters. At the same time it is shown that the common marginal likelihood (empirical Bayes) approach does not enjoy that property. (C) 2017, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.

This paper considers the input design of kernelbased regularization methods for LTI system identification. We first derive the Bayesian mean squared error matrix under the Bayesian perspective, and then use some typical scalar measures (e.g., the A-optimality, D-optimality, and E-optimality) as optimization criteria for the input design problem. Instead of directly solving the nonconvex optimization problem, we propose a two-step procedure. The first step is to solve a convex optimization and the second one is to determine the inverse image of a quadratic map. Both of these two steps can be solved efficiently by the proposed method and hence all the globally optimal inputs are found. In particular, we show that for some kernels, the optimal input under the D-optimality has an explicit expression., Funding Agencies|Thousand Youth Talents Plan - central government of China; Shenzhen Projects - Shenzhen Science and Technology Innovation Council [Ji-20170189, Ji-20160207]; Chinese University of Hong Kong, Shenzhen [PF. 01.000249, 2014.0003.23]; Swedish Research Council [2014-5894]; NSFC [61773329]

Input design is an important issue for classical system identification methods but has not been investigated for the kernel-based regularization method (KRM) until very recently. In this paper, we consider in the time domain the input design problem of KRMs for LTI system identification. Different from the recent result, we adopt a Bayesian perspective and in particular make use of scalar measures (e.g., the $A$-optimality, $D$-optimality, and $E$-optimality) of the Bayesian mean square error matrix as the design criteria subject to power-constraint on the input. Instead to solve the optimization problem directly, we propose a two-step procedure. In the first step, by making suitable assumptions on the unknown input, we construct a quadratic map (transformation) of the input such that the transformed input design problems are convex, the number of optimization variables is independent of the number of input data, and their global minima can be found efficiently by applying well-developed convex optimization software packages. In the second step, we derive the expression of the optimal input based on the global minima found in the first step by solving the inverse image of the quadratic map. In addition, we derive analytic results for some special types of fixed kernels, which provide insights on the input design and also its dependence on the kernel structure.

The kernel-based regularization method has two core issues: kernel design and hyperparameter estimation. In this paper, we focus on the second issue and study the properties of several hyperparameter estimators including the empirical Bayes (EB) estimator, two Stein's unbiased risk estimators (SURE) and their corresponding Oracle counterparts, with an emphasis on the asymptotic properties of these hyperparameter estimators. To this goal, we first derive and then rewrite the first order optimality conditions of these hyperparameter estimators, leading to several insights on these hyperparameter estimators. Then we show that as the number of data goes to infinity, the two SUREs converge to the best hyperparameter minimizing the corresponding mean square error, respectively, while the more widely used EB estimator converges to another best hyperparameter minimizing the expectation of the EB estimation criterion. This indicates that the two SUREs are asymptotically optimal but the EB estimator is not. Surprisingly, the convergence rate of two SUREs is slower than that of the EB estimator, and moreover, unlike the two SUREs, the EB estimator is independent of the convergence rate of $\Phi^T\Phi/N$ to its limit, where $\Phi$ is the regression matrix and $N$ is the number of data. A Monte Carlo simulation is provided to demonstrate the theoretical results.