x 104

Size: px
Start display at page:

Download "x 104"

Transcription

1 Departement Elektrotechniek ESAT-SISTA/TR 98-3 Identication of the circulant modulated Poisson process: a time domain approach Katrien De Cock, Tony Van Gestel and Bart De Moor 2 April 998 Submitted for publication in Proceedings of MTNS'98. This report is available by anonymous ftp from ftp.esat.kuleuven.ac.be in the directory pub/sista/decock/reports/98-3.ps.gz 2 K.U.Leuven, Dept. of Electrical Engineering (ESAT), Research group SISTA, Kardinaal Mercierlaan 94, 300 Leuven, Belgium, Tel. 32/6/ , Fax 32/6/ , WWW: E- mail: katrien.decock@esat.kuleuven.ac.be, tony.vangestel@esat.kuleuven.ac.be, bart.demoor@esat.kuleuven.ac.be. Bart De Moor is a senior Research Associate with the F.W.O. (Fund for Scientic Research-Flanders). Tony Van Gestel is a Research Assistant with the F.W.O. (Fund for Scientic Research-Flanders). Katrien De Cock is a Research Assistant with the I.W.T. (Flemish Institute for Scientic and Technological Research in Industry). Work supported by the Flemish Government ( Administration of Science and Innovation (Concerted Research Action MIPS: Model-based Information Processing Systems, Bilateral International Collaboration: Modelling and Identication of nonlinear systems, IWT-Eureka SINOPSYS: Model-based structural monitoring using in-operation system identication), FWO-Vlaanderen: Analysis and design of matrix algorithms for adaptive signal processing, system identication and control, based on concepts from continuous time system theory and dierential geometry, Numerical algorithms for subspace system identication: Extension towards specic applications, FWO-Onderzoeksgemeenschappen: Identication and Control of Complex Systems, Advanced Numerical Methods for Mathematical Modelling); Belgian Federal Government( Interuniversity Attraction Pole IUAP IV/02: Modelling, Identication, Simulation and Control of Complex Systems, Interuniversity Attraction Pole IUAP IV/24: IMechS: Intelligent Mechatronic Systems); European Commission: (Human Capital and Mobility: SCIENCE-ERNSI: European Research Network for System Identication.)

2 Identication of the circulant modulated Poisson process: a time domain approach Katrien De Cock y, Tony Van Gestel z and Bart De Moor x K.U.Leuven, ESAT-SISTA, Kardinaal Mercierlaan 94, 300 Leuven, Belgium, Tel. 32/6/ , Fax 32/6/ , katrien.decock@esat.kuleuven.ac.be, tony.vangestel@esat.kuleuven.ac.be, bart.demoor@esat.kuleuven.ac.be. Keywords: trac modelling, circulant modulated Poisson process, ATM Abstract In this paper we discuss a new time domain method for the trac identication problem in ATM networks. Our identication approach for the circulant modulated Poisson process (CMPP) consists of two steps: the identication of the rst order parameters and the determination of the circulant stochastic matrix which matches the second order statistics of the data. Introduction Li et al. [6] have indicated that mathematical models can be used to perform several tasks in control mechanisms of ATM networks. The models they propose are measurement based and include the time correlation of trac. Whereas the approach of Li et al. [6, 7, 8] is mainly based on the frequency domain, this paper is concerned with measurement based parameter estimation in the time domain: the models match the cumulative distribution function and the autocorrelation function of the measured data as described in []. This paper is concerned with the modelling of the arrival process of cells in one node of the network. The arrival process a k (k = 0; ; 2; : : :) is the aggregated number of Work supported by the Flemish Government (BOF (GOA- MIPS), AWI (Bil. Int. Coll.), FWO (projects, grants, res. comm. (ICCoS)), IWT (IWT-VCST (CVT), ITA (ISIS), EU- REKA (Sinopsys))), the Belgian Federal Government (IUAP IV- 02, IUAP IMechS), the European Commission (TMR (Alapedes), ACTS (Aspect), SCIENCE (ERNSI)). y Research Assistant with the I.W.T. (Flemish Institute for Scientic and Technological Research in Industry). z Research Assistant with the F.W.O. (Fund for Scientic Research-Flanders). x Senior Research Associate with the F.W.O. (Fund for Scientic Research-Flanders). cells that arrive per time unit. This sequence is measured and forms the input of our identication algorithm. The research of Li and Hwang [7] has demonstrated that only the rst order and second order statistics have a signicant impact on queueing performance. Therefore we will identify a model that has rst and second order statistics that approximate those of the data as well as possible. The model should also t into queueing analysis which is used to evaluate the network performance. A Markov modulated Poisson process (MMPP) satises both conditions, but a more tractable model structure is the circulant modulated Poisson process (CMPP) which is a restricted version of the MMPP. 2 The circulant modulated Poisson process In this section we rst explain the structure of a circulant modulated Poisson process (CMPP). Then we discuss some important properties of the CMPP with respect to our identication method. Because we want to approximate the rst and second order statistics of the data with those of a CMPP, we give the expressions for the rst and second order statistics of the model. 2. The model structure A CMPP of order N C consists of an N C -state circulant Markov chain in which each state i (i = ; 2; : : : ; N C ) represents a Poisson process with rate i. In this paper we only consider discrete time Markov chains. The circulant Markov chain is characterised by a circulant stochastic transition matrix Q: Q = q q 2 : : : q NC q NC q : : : q NC? ; q 2 q 3 : : : q

3 where XN C i= q i 0; for i = ; 2; : : : ; N C () q i = : (2) Throughout this paper, we will deal exclusively with the steady state case, where the state distribution is time independent: T = T Q : So is the left eigenvector of the circulant stochastic matrix Q corresponding to eigenvalue. The Poisson parameters of the CMPP are arranged in a column vector 2 R NC, where each element i (i = ; 2; : : : ; N C ) is a positive real number, representing the arrival rate of the Poisson process associated with state i. At time step k, the CMPP (Q; ) will emit a k arrivals generated by the Poisson process with the arrival rate i if the state of the Markov chain is i. 2.2 Properties of a CMPP Circulant matrices have some important properties because of their special structure (see e.g. []). A circulant stochastic matrix Q is always diagonalisable. Its eigenvalue decomposition can be computed very eciently. If Q = V V? is the eigenvalue decomposition of the circulant stochastic matrix Q, then we have: V is a Fourier matrix: p?2? V (i; j) = p exp (i? )(j? ) NC N C V? = V H ; where V H denotes the complex conjugate transpose of V. The eigenvectors of a circulant matrix are independent of the entries of that matrix, they only depend on the dimension of the matrix. The product of a Fourier matrix V and a vector x is equal to the discrete Fourier transform of that vector x. This discrete Fourier transform can be computed with a fast Fourier transform (fft). The left eigenvector of a circulant stochastic matrix Q of dimension N C corresponding to the eigenvalue is a uniform vector T C = : : : (3) N C N C N C because circulant stochastic matrices are double stochastic. As a consequence, the state distribution of a CMPP is independent of the entries of the transition matrix. The eigenvalues i of a circulant matrix Q can be computed as follows: p T T : : : NC = N C V q : : : q NC T = fft q : : : q NC : (4) The vector of eigenvalues of a circulant matrix Q is equal to the fft of a row of Q. 2.3 The rst and second order statistics of a CMPP As a description for the rst order statistics of the cell arrival process a k, we use the cumulative distribution function F d, dened as F d (l) = P rfa k lg; l 2 N: The cumulative distribution function of a CMPP is given by F C (l) = N C XN C j= e?(c)j lx i=0 i! ( C) i j : (5) It does not depend on the transition matrix Q in contrast with the cumulative distribution function of a MMPP. Therefore we can uncouple the identication of the rst order and of the second order parameters. The second order statistics of the data a k are described by the autocorrelation function R(n) = Efa k a k+n g: The autocorrelation function of a CMPP can be computed as follows: R C (n) = T?Q n?e = C T N V n V H C = N 2 C fft( T C ) diag ([ n i ]) fft( C ) ; (6) where is obtained as in (4) and fft( C ) is the complex conjugate of fft( C ). Only two fft's are needed to compute the autocorrelation function of a CMPP. 3 The identication of a CMPP In this section our new identication method for the CMPP is discussed. Because of the special structure of the CMPP, the identication of the rst and second order parameters can be uncoupled. In Figure an overview is given of the complete identication method. First, the cumulative distribution function of the data is used to nd the rst order parameters of an MMPP. Those rst order MMPP parameters are then translated to CMPP parameters taking into account the restrictions of the circulant structure. This is the rst step of our identication method and it is explained in more detail in Section 3.. The second step consists of the identication of the circulant stochastic matrix by unconstrained optimisation which is discussed in Section 3.2. In Section 3.3 an example of our identication procedure is given. 3. The identication of the rst order parameters The cumulative distribution function of the data must be approximated by the cumulative distribution function of

4 minimise k 0 F d solve min kf x d? Dxk 2 s.t. x i 0 NM M M a k compute distribution and autocorrelation translate NC C MMPP into CMPP R d min k kr d? R Ck 2 2 step : step 2: identication identication of the of the Q rst order parameters circulant transition matrix Figure : Identication of the CMPP where the cumulative distribution function is matched by solving a nonnegative least squares problem followed by a translation of MMPP parameters to CMPP parameters. The autocorrelation function is matched by unconstrained optimisation. a CMPP. This implies that F d must be reconstructed as a linear combination of cumulative distributions of Poisson processes in which each coecient is equal to N C (see (5)). Our new method for the identication of the rst order parameters consists of two parts: ) Based on the cumulative distribution function of the data a k, we determine which Poisson processes are involved in the arrival process. The distribution of these N M N C Poisson processes can then be viewed as the state distribution vector M of an MMPP, which is not uniform. The Poisson parameters form the MMPP parameter M. In this rst part we approximate the cumulative distribution function of the data with a nonnegative linear combination of cumulative Poisson distributions (which in fact is the cumulative distribution of an MMPP): F d D ; where (F d ) l = F d (l? ) and F d 2 R L+ (L is the maximum number of arrivals of the given set of data). Each column D j of D represents a possible state or possible Poisson cumulative distribution. In fact, the domain of possible i 's is discretized in order to get a broad choice of candidate states. The discretisation step is not chosen to be constant but it increases linearly because the variance of a Poisson process is equal to its rate. In order to determine which of the proposed cumulative Poisson distributions take part in F d we solve min x kf d? Dxk 2 ; subject to x i 0 : We have used the algorithm of Lawson and Hanson [5] to solve this problem. Let x be the optimal value of x. The number of non-zero components of x gives the MMPP model order N M. The indices of the nonzero components of x give the ( M ) i 's. The ( M ) i 's are given by the values of the non-zero components of x. For more details we refer to [2, 4]. 2) In order to obtain a uniform distribution of the Poisson processes (see (3)) we take as many copies of each state as needed to approximate each component ( M ) i of the non-uniform distribution vector M with multiples of NC. The total number of states is then equal to N C. In Figure we refer to this second part as the translation of MMPP parameters into CMPP parameters. More details are found in [2, 3]. By imposing a certain model order N C we introduce a modelling error. However, we can determine a bound on this error as a function of N C, so that we can choose the minimal model order that still guarantees a given accuracy of the cumulative distribution function. 3.2 The identication of the circulant transition matrix For the identication of the circulant transition matrix Q we solve the following optimisation problem: min Q kr d? R C k 2 2 ; where R d is the autocorrelation function of the data and R C is the autocorrelation function of the CMPP. In order to obtain an unconstrained optimisation problem we use the parameterisation of Q, proposed by Yi and De Moor []: 8 >< >: q i = q NC = P ki 2 N + C? j= kj 2 ; for i = ; : : : ; N C? P N + C? j= kj 2 : This parameterisation takes the constraints on the circulant stochastic matrix into account (see () and (2)). The only variable in the optimisation problem is now k 2 R NC?. During the optimisation we keep the Poisson parameters C which are found in the rst step, constant. The circulant structure is exploited during the evaluation of the autocorrelation function (see (6)). Because of the ef- cient computation of the eigenvalue decomposition, the complexity of this optimisation step is O(N C log(n C )) per function evaluation.

5 R F placements x Figure 2: The cumulative distribution function (top) and the autocorrelation function (bottom) of the ethernet data (full line) and of the identied CMPP (dash-dotted line). 3.3 Example We give an example of our identication procedure where a CMPP is identied starting from measured ethernet data. The data are measurements of one hour of internet trac between the Lawrence Berkeley laboratory and the rest of the world. The traces were made by Vern Paxson [0] and they are available from the Internet Traf- c Archive of the Lawrence Berkeley National Laboratory [9]. From the lbl-pkt database we have used the measurements lbl-pkt-4. The data give the number of bytes that was sent per microsecond. We have used intervals of 0.25 seconds in order to avoid too many zeros in the data sequence. All bytes that were sent in an interval [t i ; t i + 0:25s] are grouped. Our specic dataset is available from our database for identication DAISY. It consists of data points. The order of the identied CMPP was chosen to be equal to 28. In Figure 2 we compare the cumulative distribution function (top picture) and the autocorrelation function (bottom picture) of the data with those of the identied model. The full line represents the statistics of the data and the dashdotted line gives the statistics of the CMPP. 4 Conclusions In this paper we have presented a new time domain identication algorithm for ATM network trac. The algorithm identies the circulant modulated Poisson process (CMPP). Only the rst order and second order statistics are matched, respectively the cumulative distribution function and the autocorrelation function. We identify the CMPP in two steps. The rst order l n statistics are identied very eciently by solving a nonnegative least squares problem followed by a translation of the parameters to the restricted CMPP format. The circulant stochastic transition matrix is identied by unconstrained optimisation. Due to lack of space we could not include the complete analysis of the computational complexity of our CMPP identication method. The interested reader is referred to [2]. References [] A. Berman en R.J. Plemmons. Nonnegative Matrices in the Mathematical Sciences. Classics in Applied Mathematics series, vol. 9. SIAM, Philadelphia, 994. [2] K. De Cock, T. Van Gestel and B. De Moor. Stochastic system identication for ATM network trac models: a time domain approach. Internal report ESAT-SISTA/ , K.U.Leuven, Belgium, 997. [3] K. De Cock and B. De Moor. Identication of the rst order parameters of a circulant modulated Poisson process. Internal report ESAT-SISTA/ , K.U.Leuven, Belgium, 997. Accepted for publication in Proceedings of International Conference on Telecommunications (ICT'98). [4] K. De Cock and B. De Moor. Stochastic system identication for ATM network trac models: a time domain approach. Internal report ESAT-SISTA/ , K.U.Leuven, Belgium, 998. Accepted for publication in Proceedings of EUSIPCO-98 [5] C.L. Lawson and R.J. Hanson. Solving Least Squares Problems. Classics in Applied Mathematics series, vol. 5. SIAM, Philadelphia, 995. [6] S.Q. Li and C.L. Hwang. Queue response to input correlation functions: discrete spectral analysis. IEEE/ACM Transactions on Networking, (5):522{533, 993. [7] S.Q. Li and C.L. Hwang. Queue response to input correlation functions: continuous spectral analysis.ieee/acm Transactions on Networking, (6):678{ 692, 993. [8] S.Q. Li and C.L. Hwang. On the convergence of traf- c measurement and queueing analysis: a statisticalmatching and queueing (SMAQ) Tool. IEEE/ACM Transactions on Networking, 5():95{0, 997. [9] The internet trac archive, [0] V. Paxson en S. Floyd. The failure of Poisson modeling. IEEE/ACM Transactions on Networking 3(3):226{244, 995. [] C. Yi and B. De Moor. Trac identication of ATM networks with optimization algorithms. Proceedings of the 35th IEEE Conference on Decision and Control, Kobe, Japan, pp. 277{282, This report is available by anonymous ftp from ftp.esat.kuleuven.ac.be in the directory pub/sista/decock/reports/

Departement Elektrotechniek ESAT-SISTA/TR 98- Stochastic System Identication for ATM Network Trac Models: a Time Domain Approach Katrien De Cock and Bart De Moor April 998 Accepted for publication in roceedings

More information

Optimal trac light control for a single intersection Bart De Schutter and Bart De Moor y ESAT-SISTA, KU Leuven, Kardinaal Mercierlaan 94, B-3 Leuven (

Optimal trac light control for a single intersection Bart De Schutter and Bart De Moor y ESAT-SISTA, KU Leuven, Kardinaal Mercierlaan 94, B-3 Leuven ( Katholieke Universiteit Leuven Departement Elektrotechniek ESAT-SISTA/TR 97- Optimal trac light control for a single intersection Bart De Schutter and Bart De Moor Proceedings of the 997 International

More information

The Linear Dynamic Complementarity Problem is a special case of the Extended Linear Complementarity Problem B. De Schutter 1 B. De Moor ESAT-SISTA, K.

The Linear Dynamic Complementarity Problem is a special case of the Extended Linear Complementarity Problem B. De Schutter 1 B. De Moor ESAT-SISTA, K. Katholieke Universiteit Leuven Departement Elektrotechniek ESAT-SISTA/TR 9-1 The Linear Dynamic Complementarity Problem is a special case of the Extended Linear Complementarity Problem 1 Bart De Schutter

More information

On max-algebraic models for transportation networks

On max-algebraic models for transportation networks K.U.Leuven Department of Electrical Engineering (ESAT) SISTA Technical report 98-00 On max-algebraic models for transportation networks R. de Vries, B. De Schutter, and B. De Moor If you want to cite this

More information

Block-row Hankel Weighted Low Rank Approximation 1

Block-row Hankel Weighted Low Rank Approximation 1 Katholieke Universiteit Leuven Departement Elektrotechniek ESAT-SISTA/TR 03-105 Block-row Hankel Weighted Low Rank Approximation 1 Mieke Schuermans, Philippe Lemmerling and Sabine Van Huffel 2 July 2003

More information

Optimal traffic light control for a single intersection

Optimal traffic light control for a single intersection KULeuven Department of Electrical Engineering (ESAT) SISTA Technical report 97- Optimal traffic light control for a single intersection B De Schutter and B De Moor If you want to cite this report, please

More information

Departement Elektrotechniek ESAT-SISTA/TR Dynamical System Prediction: a Lie algebraic approach for a novel. neural architecture 1

Departement Elektrotechniek ESAT-SISTA/TR Dynamical System Prediction: a Lie algebraic approach for a novel. neural architecture 1 Katholieke Universiteit Leuven Departement Elektrotechniek ESAT-SISTA/TR 1995-47 Dynamical System Prediction: a Lie algebraic approach for a novel neural architecture 1 Yves Moreau and Joos Vandewalle

More information

Structured weighted low rank approximation 1

Structured weighted low rank approximation 1 Departement Elektrotechniek ESAT-SISTA/TR 03-04 Structured weighted low rank approximation 1 Mieke Schuermans, Philippe Lemmerling and Sabine Van Huffel 2 January 2003 Accepted for publication in Numerical

More information

Matrix factorization and minimal state space realization in the max-plus algebra

Matrix factorization and minimal state space realization in the max-plus algebra KULeuven Department of Electrical Engineering (ESAT) SISTA Technical report 96-69 Matrix factorization and minimal state space realization in the max-plus algebra B De Schutter and B De Moor If you want

More information

SVD-based optimal ltering with applications to noise reduction in speech signals Simon Doclo ESAT - SISTA, Katholieke Universiteit Leuven Kardinaal Me

SVD-based optimal ltering with applications to noise reduction in speech signals Simon Doclo ESAT - SISTA, Katholieke Universiteit Leuven Kardinaal Me Departement Elektrotechniek ESAT-SISTA/TR 999- SVD-based Optimal Filtering with Applications to Noise Reduction in Speech Signals Simon Doclo, Marc Moonen April, 999 Internal report This report is available

More information

The extended linear complementarity problem and its applications in the analysis and control of discrete event systems and hybrid systems

The extended linear complementarity problem and its applications in the analysis and control of discrete event systems and hybrid systems KULeuven Department of Electrical Engineering (ESAT) SISTA Technical report 97-34 The extended linear complementarity problem and its applications in the analysis and control of discrete event systems

More information

The QR decomposition and the singular value decomposition in the symmetrized max-plus algebra

The QR decomposition and the singular value decomposition in the symmetrized max-plus algebra K.U.Leuven Department of Electrical Engineering (ESAT) SISTA Technical report 96-70 The QR decomposition and the singular value decomposition in the symmetrized max-plus algebra B. De Schutter and B. De

More information

State space transformations and state space realization in the max algebra

State space transformations and state space realization in the max algebra K.U.Leuven Department of Electrical Engineering (ESAT) SISTA Technical report 95-10 State space transformations and state space realization in the max algebra B. De Schutter and B. De Moor If you want

More information

Information, Covariance and Square-Root Filtering in the Presence of Unknown Inputs 1

Information, Covariance and Square-Root Filtering in the Presence of Unknown Inputs 1 Katholiee Universiteit Leuven Departement Eletrotechnie ESAT-SISTA/TR 06-156 Information, Covariance and Square-Root Filtering in the Presence of Unnown Inputs 1 Steven Gillijns and Bart De Moor 2 October

More information

Departement Elektrotechniek ESAT-SISTA/TR About the choice of State Space Basis in Combined. Deterministic-Stochastic Subspace Identication 1

Departement Elektrotechniek ESAT-SISTA/TR About the choice of State Space Basis in Combined. Deterministic-Stochastic Subspace Identication 1 Katholieke Universiteit Leuven Departement Elektrotechniek ESAT-SISTA/TR 994-24 About the choice of State Space asis in Combined Deterministic-Stochastic Subspace Identication Peter Van Overschee and art

More information

Maximum Likelihood Estimation and Polynomial System Solving

Maximum Likelihood Estimation and Polynomial System Solving Maximum Likelihood Estimation and Polynomial System Solving Kim Batselier Philippe Dreesen Bart De Moor Department of Electrical Engineering (ESAT), SCD, Katholieke Universiteit Leuven /IBBT-KULeuven Future

More information

Linear dynamic filtering with noisy input and output 1

Linear dynamic filtering with noisy input and output 1 Katholieke Universiteit Leuven Departement Elektrotechniek ESAT-SISTA/TR 22-191 Linear dynamic filtering with noisy input and output 1 Ivan Markovsky and Bart De Moor 2 2 November 22 Published in the proc

More information

A note on the characteristic equation in the max-plus algebra

A note on the characteristic equation in the max-plus algebra K.U.Leuven Department of Electrical Engineering (ESAT) SISTA Technical report 96-30 A note on the characteristic equation in the max-plus algebra B. De Schutter and B. De Moor If you want to cite this

More information

Master-Slave Synchronization using. Dynamic Output Feedback. Kardinaal Mercierlaan 94, B-3001 Leuven (Heverlee), Belgium

Master-Slave Synchronization using. Dynamic Output Feedback. Kardinaal Mercierlaan 94, B-3001 Leuven (Heverlee), Belgium Master-Slave Synchronization using Dynamic Output Feedback J.A.K. Suykens 1, P.F. Curran and L.O. Chua 1 Katholieke Universiteit Leuven, Dept. of Electr. Eng., ESAT-SISTA Kardinaal Mercierlaan 94, B-1

More information

Compactly supported RBF kernels for sparsifying the Gram matrix in LS-SVM regression models

Compactly supported RBF kernels for sparsifying the Gram matrix in LS-SVM regression models Compactly supported RBF kernels for sparsifying the Gram matrix in LS-SVM regression models B. Hamers, J.A.K. Suykens, B. De Moor K.U.Leuven, ESAT-SCD/SISTA, Kasteelpark Arenberg, B-3 Leuven, Belgium {bart.hamers,johan.suykens}@esat.kuleuven.ac.be

More information

Data Assimilation in 2D Magneto-Hydrodynamics Systems 1

Data Assimilation in 2D Magneto-Hydrodynamics Systems 1 Katholieke Universiteit Leuven Departement Elektrotechniek ESAT-SISTA/TR 0-7 Data Assimilation in D Magneto-Hydrodynamics Systems Oscar Barrero Mendoza, Dennis S. Bernstein, and Bart L.R. De Moor. September

More information

Embedding Recurrent Neural Networks into Predator-Prey Models Yves Moreau, St phane Louies 2, Joos Vandewalle, L on renig 2 Katholieke Universiteit Le

Embedding Recurrent Neural Networks into Predator-Prey Models Yves Moreau, St phane Louies 2, Joos Vandewalle, L on renig 2 Katholieke Universiteit Le Katholieke Universiteit Leuven Departement Elektrotechniek ESAT-SISTA/TR 998-2 Embedding Recurrent Neural Networks into Predator-Prey Models Yves Moreau 2, Stephane Louies 3, Joos Vandewalle 2, Leon renig

More information

The characteristic equation and minimal state space realization of SISO systems in the max algebra

The characteristic equation and minimal state space realization of SISO systems in the max algebra KULeuven Department of Electrical Engineering (ESAT) SISTA Technical report 93-57a The characteristic equation and minimal state space realization of SISO systems in the max algebra B De Schutter and B

More information

Packet Size

Packet Size Long Range Dependence in vbns ATM Cell Level Trac Ronn Ritke y and Mario Gerla UCLA { Computer Science Department, 405 Hilgard Ave., Los Angeles, CA 90024 ritke@cs.ucla.edu, gerla@cs.ucla.edu Abstract

More information

MANY studies indicate the importance of high burstiness

MANY studies indicate the importance of high burstiness 612 IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 16, NO. 5, JUNE 1998 Fast Algorithms for Measurement-Based Traffic Modeling Hao Che and San-qi Li Abstract This paper develops fast algorithms

More information

Joint Regression and Linear Combination of Time Series for Optimal Prediction

Joint Regression and Linear Combination of Time Series for Optimal Prediction Joint Regression and Linear Combination of Time Series for Optimal Prediction Dries Geebelen 1, Kim Batselier 1, Philippe Dreesen 1, Marco Signoretto 1, Johan Suykens 1, Bart De Moor 1, Joos Vandewalle

More information

Affine iterations on nonnegative vectors

Affine iterations on nonnegative vectors Affine iterations on nonnegative vectors V. Blondel L. Ninove P. Van Dooren CESAME Université catholique de Louvain Av. G. Lemaître 4 B-348 Louvain-la-Neuve Belgium Introduction In this paper we consider

More information

Subspace angles and distances between ARMA models

Subspace angles and distances between ARMA models ÔÖØÑÒØ ÐØÖÓØÒ Ë̹ËÁËÌ»ÌÊ ß ËÙ Ô ÒÐ Ò ØÒ ØÛÒ ÊÅ ÑÓÐ ÃØÖÒ Ó Ò ÖØ ÅÓÓÖ ÅÖ ÈÙÐ Ò ÈÖÓÒ Ó Ø ÓÙÖØÒØ ÁÒØÖÒØÓÒÐ ËÝÑÔÓ ÙÑ Ó ÅØÑØÐ ÌÓÖÝ Ó ÆØÛÓÖ Ò ËÝ ØÑ ÅÌÆË µ ÈÖÔÒÒ ÖÒ ÂÙÒ ß Ì ÖÔÓÖØ ÚÐÐ Ý ÒÓÒÝÑÓÙ ØÔ ÖÓÑ ØÔº غÙÐÙÚÒºº

More information

arxiv: v1 [math.ag] 10 Oct 2011

arxiv: v1 [math.ag] 10 Oct 2011 Are diverging CP components always nearly proportional? Alwin Stegeman and Lieven De Lathauwer November, 8 Abstract arxiv:.988v [math.ag] Oct Fitting a Candecomp/Parafac (CP) decomposition (also nown as

More information

CAC investigation for video and data

CAC investigation for video and data CAC investigation for video and data E.Aarstad a, S.Blaabjerg b, F.Cerdan c, S.Peeters d and K.Spaey d a Telenor Research & Development, P.O. Box 8, N-7 Kjeller, Norway,egil.aarstad@fou.telenor.no b Tele

More information

On Weighted Structured Total Least Squares

On Weighted Structured Total Least Squares On Weighted Structured Total Least Squares Ivan Markovsky and Sabine Van Huffel KU Leuven, ESAT-SCD, Kasteelpark Arenberg 10, B-3001 Leuven, Belgium {ivanmarkovsky, sabinevanhuffel}@esatkuleuvenacbe wwwesatkuleuvenacbe/~imarkovs

More information

Math 1553, Introduction to Linear Algebra

Math 1553, Introduction to Linear Algebra Learning goals articulate what students are expected to be able to do in a course that can be measured. This course has course-level learning goals that pertain to the entire course, and section-level

More information

communication networks

communication networks Positive matrices associated with synchronised communication networks Abraham Berman Department of Mathematics Robert Shorten Hamilton Institute Douglas Leith Hamilton Instiute The Technion NUI Maynooth

More information

Queue Response to Input Correlation Functions: Continuous Spectral Analysis. University of Texas at Austin. Austin, Texas

Queue Response to Input Correlation Functions: Continuous Spectral Analysis. University of Texas at Austin. Austin, Texas Queue Response to Input Correlation Functions: Continuous Spectral Analysis San-i Li Chia-Lin Hwang Department of Electrical and Computer Engineering University of Texas at Austin Austin, Texas 78712 August

More information

Departement Elektrotechniek ESAT-SISTA/TR Minimal state space realization of SISO systems in the max algebra 1.

Departement Elektrotechniek ESAT-SISTA/TR Minimal state space realization of SISO systems in the max algebra 1. Katholieke Universiteit Leuven Departement Elektrotechniek ESAT-SISTA/TR 9- inimal state space realization of SISO systems in the max algebra 1 Bart De Schutter and Bart De oor October 199 A short version

More information

Minimal realization in the max algebra is an extended linear complementarity problem

Minimal realization in the max algebra is an extended linear complementarity problem K.U.Leuven Department of Electrical Engineering (ESAT) SISTA Technical report 93-70 Minimal realization in the max algebra is an extended linear complementarity problem B. De Schutter and B. De Moor December

More information

TMA Calculus 3. Lecture 21, April 3. Toke Meier Carlsen Norwegian University of Science and Technology Spring 2013

TMA Calculus 3. Lecture 21, April 3. Toke Meier Carlsen Norwegian University of Science and Technology Spring 2013 TMA4115 - Calculus 3 Lecture 21, April 3 Toke Meier Carlsen Norwegian University of Science and Technology Spring 2013 www.ntnu.no TMA4115 - Calculus 3, Lecture 21 Review of last week s lecture Last week

More information

Multiplicative Multifractal Modeling of. Long-Range-Dependent (LRD) Trac in. Computer Communications Networks. Jianbo Gao and Izhak Rubin

Multiplicative Multifractal Modeling of. Long-Range-Dependent (LRD) Trac in. Computer Communications Networks. Jianbo Gao and Izhak Rubin Multiplicative Multifractal Modeling of Long-Range-Dependent (LRD) Trac in Computer Communications Networks Jianbo Gao and Izhak Rubin Electrical Engineering Department, University of California, Los Angeles

More information

System occupancy of a two-class batch-service queue with class-dependent variable server capacity

System occupancy of a two-class batch-service queue with class-dependent variable server capacity System occupancy of a two-class batch-service queue with class-dependent variable server capacity Jens Baetens 1, Bart Steyaert 1, Dieter Claeys 1,2, and Herwig Bruneel 1 1 SMACS Research Group, Dept.

More information

SEPARATION OF ACOUSTIC SIGNALS USING SELF-ORGANIZING NEURAL NETWORKS. Temujin Gautama & Marc M. Van Hulle

SEPARATION OF ACOUSTIC SIGNALS USING SELF-ORGANIZING NEURAL NETWORKS. Temujin Gautama & Marc M. Van Hulle SEPARATION OF ACOUSTIC SIGNALS USING SELF-ORGANIZING NEURAL NETWORKS Temujin Gautama & Marc M. Van Hulle K.U.Leuven, Laboratorium voor Neuro- en Psychofysiologie Campus Gasthuisberg, Herestraat 49, B-3000

More information

In: Proc. BENELEARN-98, 8th Belgian-Dutch Conference on Machine Learning, pp 9-46, 998 Linear Quadratic Regulation using Reinforcement Learning Stephan ten Hagen? and Ben Krose Department of Mathematics,

More information

ON THE REGULARIZATION OF CANONICAL CORRELATION ANALYSIS. Tijl De Bie. Bart De Moor

ON THE REGULARIZATION OF CANONICAL CORRELATION ANALYSIS. Tijl De Bie. Bart De Moor ON THE REGULARIZATION OF CANONICAL CORRELATION ANALYSIS Tijl De Bie Katholiee Universiteit Leuven ESAT-SCD Kasteelpar Arenberg 0 300 Leuven tijl.debie@esat.uleuven.ac.be Bart De Moor Katholiee Universiteit

More information

Abstract In this paper we propose and motivate the emerging technique of Independent Component Analysis, also known as Blind Source Separation, as an

Abstract In this paper we propose and motivate the emerging technique of Independent Component Analysis, also known as Blind Source Separation, as an Departement Elektrotechniek ESAT-SISTA/TR 1998-127 Fetal Electrocardiogram Extraction by Blind Source Subspace Separation 1 Lieven De Lathauwer, Bart De Moor and Joos Vandewalle 2 October 1999 Extended

More information

Linear Algebra review Powers of a diagonalizable matrix Spectral decomposition

Linear Algebra review Powers of a diagonalizable matrix Spectral decomposition Linear Algebra review Powers of a diagonalizable matrix Spectral decomposition Prof. Tesler Math 283 Fall 2016 Also see the separate version of this with Matlab and R commands. Prof. Tesler Diagonalizing

More information

Institute for Advanced Computer Studies. Department of Computer Science. On Markov Chains with Sluggish Transients. G. W. Stewart y.

Institute for Advanced Computer Studies. Department of Computer Science. On Markov Chains with Sluggish Transients. G. W. Stewart y. University of Maryland Institute for Advanced Computer Studies Department of Computer Science College Park TR{94{77 TR{3306 On Markov Chains with Sluggish Transients G. W. Stewart y June, 994 ABSTRACT

More information

NL q theory: checking and imposing stability of recurrent neural networks for nonlinear modelling J.A.K. Suykens, J. Vandewalle, B. De Moor Katholieke

NL q theory: checking and imposing stability of recurrent neural networks for nonlinear modelling J.A.K. Suykens, J. Vandewalle, B. De Moor Katholieke NL q theory: checing and imposing stability of recurrent neural networs for nonlinear modelling J.A.K. Suyens, J. Vandewalle, B. De Moor Katholiee Universiteit Leuven, Dept. of Electr. Eng., ESAT-SISTA

More information

UMIACS-TR July CS-TR 2721 Revised March Perturbation Theory for. Rectangular Matrix Pencils. G. W. Stewart.

UMIACS-TR July CS-TR 2721 Revised March Perturbation Theory for. Rectangular Matrix Pencils. G. W. Stewart. UMIAS-TR-9-5 July 99 S-TR 272 Revised March 993 Perturbation Theory for Rectangular Matrix Pencils G. W. Stewart abstract The theory of eigenvalues and eigenvectors of rectangular matrix pencils is complicated

More information

Math 307 Learning Goals. March 23, 2010

Math 307 Learning Goals. March 23, 2010 Math 307 Learning Goals March 23, 2010 Course Description The course presents core concepts of linear algebra by focusing on applications in Science and Engineering. Examples of applications from recent

More information

ELECTRONIC CONTROL OF CONTINUOUSLY VARIABLE TRANS- The technology of a Continuously Variable Transmission (CVT) has been

ELECTRONIC CONTROL OF CONTINUOUSLY VARIABLE TRANS- The technology of a Continuously Variable Transmission (CVT) has been 1 ELECTRONIC CONTROL OF CONTINUOUSLY VARIABLE TRANS- MISSIONS. Paul Vanvuchelen, Christiaan Moons, Willem Minten, Bart De Moor ESAT - Katholieke Universiteit Leuven, Kardinaal Mercierlaan 94, 31 Leuven

More information

Financial Time Series Prediction Using Least Squares Support Vector Machines Within the Evidence Framework

Financial Time Series Prediction Using Least Squares Support Vector Machines Within the Evidence Framework IEEE TRANSACTIONS ON NEURAL NETWORKS, VOL. 12, NO. 4, JULY 2001 809 Financial Time Series Prediction Using Least Squares Support Vector Machines Within the Evidence Framework Tony Van Gestel, Johan A.

More information

Defense Technical Information Center Compilation Part Notice

Defense Technical Information Center Compilation Part Notice UNCLASSIFIED Defense Technical Information Center Compilation Part Notice ADP012911 TITLE: Low Temperature Scanning Tunneling Spectroscopy of Different Individual Impurities on GaAs [110] Surface and in

More information

N.G.Bean, D.A.Green and P.G.Taylor. University of Adelaide. Adelaide. Abstract. process of an MMPP/M/1 queue is not a MAP unless the queue is a

N.G.Bean, D.A.Green and P.G.Taylor. University of Adelaide. Adelaide. Abstract. process of an MMPP/M/1 queue is not a MAP unless the queue is a WHEN IS A MAP POISSON N.G.Bean, D.A.Green and P.G.Taylor Department of Applied Mathematics University of Adelaide Adelaide 55 Abstract In a recent paper, Olivier and Walrand (994) claimed that the departure

More information

Linearly-solvable Markov decision problems

Linearly-solvable Markov decision problems Advances in Neural Information Processing Systems 2 Linearly-solvable Markov decision problems Emanuel Todorov Department of Cognitive Science University of California San Diego todorov@cogsci.ucsd.edu

More information

Damage Assessment of the Z24 bridge by FE Model Updating. Anne Teughels 1, Guido De Roeck

Damage Assessment of the Z24 bridge by FE Model Updating. Anne Teughels 1, Guido De Roeck Damage Assessment of the Z24 bridge by FE Model Updating Anne Teughels, Guido De Roeck Katholieke Universiteit Leuven, Department of Civil Engineering Kasteelpark Arenberg 4, B 3 Heverlee, Belgium Anne.Teughels@bwk.kuleuven.ac.be

More information

Linear Algebra review Powers of a diagonalizable matrix Spectral decomposition

Linear Algebra review Powers of a diagonalizable matrix Spectral decomposition Linear Algebra review Powers of a diagonalizable matrix Spectral decomposition Prof. Tesler Math 283 Fall 2018 Also see the separate version of this with Matlab and R commands. Prof. Tesler Diagonalizing

More information

Equivalence of state representations for hidden Markov models

Equivalence of state representations for hidden Markov models Systems & Control Letters 57 (2008) 410 419 www.elsevier.com/locate/sysconle Equivalence of state representations for hidden Markov models Bart Vanluyten, Jan C. Willems, Bart De Moor K.U.Leuven, ESAT/SCD(SISTA),

More information

April 26, Applied mathematics PhD candidate, physics MA UC Berkeley. Lecture 4/26/2013. Jed Duersch. Spd matrices. Cholesky decomposition

April 26, Applied mathematics PhD candidate, physics MA UC Berkeley. Lecture 4/26/2013. Jed Duersch. Spd matrices. Cholesky decomposition Applied mathematics PhD candidate, physics MA UC Berkeley April 26, 2013 UCB 1/19 Symmetric positive-definite I Definition A symmetric matrix A R n n is positive definite iff x T Ax > 0 holds x 0 R n.

More information

5 Eigenvalues and Diagonalization

5 Eigenvalues and Diagonalization Linear Algebra (part 5): Eigenvalues and Diagonalization (by Evan Dummit, 27, v 5) Contents 5 Eigenvalues and Diagonalization 5 Eigenvalues, Eigenvectors, and The Characteristic Polynomial 5 Eigenvalues

More information

[4] T. I. Seidman, \\First Come First Serve" is Unstable!," tech. rep., University of Maryland Baltimore County, 1993.

[4] T. I. Seidman, \\First Come First Serve is Unstable!, tech. rep., University of Maryland Baltimore County, 1993. [2] C. J. Chase and P. J. Ramadge, \On real-time scheduling policies for exible manufacturing systems," IEEE Trans. Automat. Control, vol. AC-37, pp. 491{496, April 1992. [3] S. H. Lu and P. R. Kumar,

More information

Math 307 Learning Goals

Math 307 Learning Goals Math 307 Learning Goals May 14, 2018 Chapter 1 Linear Equations 1.1 Solving Linear Equations Write a system of linear equations using matrix notation. Use Gaussian elimination to bring a system of linear

More information

THE PROBLEMS OF ROBUST LPC PARAMETRIZATION FOR. Petr Pollak & Pavel Sovka. Czech Technical University of Prague

THE PROBLEMS OF ROBUST LPC PARAMETRIZATION FOR. Petr Pollak & Pavel Sovka. Czech Technical University of Prague THE PROBLEMS OF ROBUST LPC PARAMETRIZATION FOR SPEECH CODING Petr Polla & Pavel Sova Czech Technical University of Prague CVUT FEL K, 66 7 Praha 6, Czech Republic E-mail: polla@noel.feld.cvut.cz Abstract

More information

One important issue in the study of queueing systems is to characterize departure processes. Study on departure processes was rst initiated by Burke (

One important issue in the study of queueing systems is to characterize departure processes. Study on departure processes was rst initiated by Burke ( The Departure Process of the GI/G/ Queue and Its MacLaurin Series Jian-Qiang Hu Department of Manufacturing Engineering Boston University 5 St. Mary's Street Brookline, MA 2446 Email: hqiang@bu.edu June

More information

Baltzer Journals Received 7 January 1997; revised 16 April 1997 An Interpolation Approximation for the GI/G/1 Queue Based on Multipoint Pade Approxima

Baltzer Journals Received 7 January 1997; revised 16 April 1997 An Interpolation Approximation for the GI/G/1 Queue Based on Multipoint Pade Approxima Baltzer Journals Received 7 January 1997; revised 16 April 1997 An Interpolation Approximation for the GI/G/1 Queue Based on Multipoint Pade Approximation Muckai K Girish 1 and Jian-Qiang Hu 1 Telesis

More information

Approximation of the Karhunen}Loève transformation and its application to colour images

Approximation of the Karhunen}Loève transformation and its application to colour images Signal Processing: Image Communication 6 (00) 54}55 Approximation of the Karhunen}Loève transformation and its application to colour images ReH mi Kouassi, Pierre Gouton*, Michel Paindavoine Laboratoire

More information

VII Selected Topics. 28 Matrix Operations

VII Selected Topics. 28 Matrix Operations VII Selected Topics Matrix Operations Linear Programming Number Theoretic Algorithms Polynomials and the FFT Approximation Algorithms 28 Matrix Operations We focus on how to multiply matrices and solve

More information

Data assimilation for magnetohydrodynamics systems

Data assimilation for magnetohydrodynamics systems Journal of Computational and Applied Mathematics 89 (00) 9 www.elsevier.com/locate/cam Data assimilation for magnetohydrodynamics systems O. Barrero Mendoza a,, B. De Moor a, D.S. Bernstein b a Department

More information

SYSTEM RECONSTRUCTION FROM SELECTED HOS REGIONS. Haralambos Pozidis and Athina P. Petropulu. Drexel University, Philadelphia, PA 19104

SYSTEM RECONSTRUCTION FROM SELECTED HOS REGIONS. Haralambos Pozidis and Athina P. Petropulu. Drexel University, Philadelphia, PA 19104 SYSTEM RECOSTRUCTIO FROM SELECTED HOS REGIOS Haralambos Pozidis and Athina P. Petropulu Electrical and Computer Engineering Department Drexel University, Philadelphia, PA 94 Tel. (25) 895-2358 Fax. (25)

More information

Transductively Learning from Positive Examples Only

Transductively Learning from Positive Examples Only Transductively Learning from Positive Examples Only Kristiaan Pelckmans and Johan A.K. Suykens K.U.Leuven - ESAT - SCD/SISTA, Kasteelpark Arenberg 10, B-3001 Leuven, Belgium Abstract. This paper considers

More information

CS 246 Review of Linear Algebra 01/17/19

CS 246 Review of Linear Algebra 01/17/19 1 Linear algebra In this section we will discuss vectors and matrices. We denote the (i, j)th entry of a matrix A as A ij, and the ith entry of a vector as v i. 1.1 Vectors and vector operations A vector

More information

Eigenvalue problems and optimization

Eigenvalue problems and optimization Notes for 2016-04-27 Seeking structure For the past three weeks, we have discussed rather general-purpose optimization methods for nonlinear equation solving and optimization. In practice, of course, we

More information

A NOVEL APPROACH TO THE ESTIMATION OF THE HURST PARAMETER IN SELF-SIMILAR TRAFFIC

A NOVEL APPROACH TO THE ESTIMATION OF THE HURST PARAMETER IN SELF-SIMILAR TRAFFIC Proceedings of IEEE Conference on Local Computer Networks, Tampa, Florida, November 2002 A NOVEL APPROACH TO THE ESTIMATION OF THE HURST PARAMETER IN SELF-SIMILAR TRAFFIC Houssain Kettani and John A. Gubner

More information

Advancing density waves and phase transitions in a velocity dependent randomization traffic cellular automaton

Advancing density waves and phase transitions in a velocity dependent randomization traffic cellular automaton Katholieke Universiteit Leuven Departement Elektrotechniek ESAT-SCD (SISTA) / TR 3- Advancing density waves and phase transitions in a velocity dependent randomization traffic cellular automaton Sven Maerivoet

More information

Fast Linear Iterations for Distributed Averaging 1

Fast Linear Iterations for Distributed Averaging 1 Fast Linear Iterations for Distributed Averaging 1 Lin Xiao Stephen Boyd Information Systems Laboratory, Stanford University Stanford, CA 943-91 lxiao@stanford.edu, boyd@stanford.edu Abstract We consider

More information

October 7, :8 WSPC/WS-IJWMIP paper. Polynomial functions are renable

October 7, :8 WSPC/WS-IJWMIP paper. Polynomial functions are renable International Journal of Wavelets, Multiresolution and Information Processing c World Scientic Publishing Company Polynomial functions are renable Henning Thielemann Institut für Informatik Martin-Luther-Universität

More information

Performance Comparison of Two Implementations of the Leaky. LMS Adaptive Filter. Scott C. Douglas. University of Utah. Salt Lake City, Utah 84112

Performance Comparison of Two Implementations of the Leaky. LMS Adaptive Filter. Scott C. Douglas. University of Utah. Salt Lake City, Utah 84112 Performance Comparison of Two Implementations of the Leaky LMS Adaptive Filter Scott C. Douglas Department of Electrical Engineering University of Utah Salt Lake City, Utah 8411 Abstract{ The leaky LMS

More information

Quadratic and Copositive Lyapunov Functions and the Stability of Positive Switched Linear Systems

Quadratic and Copositive Lyapunov Functions and the Stability of Positive Switched Linear Systems Proceedings of the 2007 American Control Conference Marriott Marquis Hotel at Times Square New York City, USA, July 11-13, 2007 WeA20.1 Quadratic and Copositive Lyapunov Functions and the Stability of

More information

Math Camp Notes: Linear Algebra II

Math Camp Notes: Linear Algebra II Math Camp Notes: Linear Algebra II Eigenvalues Let A be a square matrix. An eigenvalue is a number λ which when subtracted from the diagonal elements of the matrix A creates a singular matrix. In other

More information

Markov chains and the number of occurrences of a word in a sequence ( , 11.1,2,4,6)

Markov chains and the number of occurrences of a word in a sequence ( , 11.1,2,4,6) Markov chains and the number of occurrences of a word in a sequence (4.5 4.9,.,2,4,6) Prof. Tesler Math 283 Fall 208 Prof. Tesler Markov Chains Math 283 / Fall 208 / 44 Locating overlapping occurrences

More information

Characterization and Modeling of Long-Range Dependent Telecommunication Traffic

Characterization and Modeling of Long-Range Dependent Telecommunication Traffic -- -- Characterization and Modeling of Long-Range Dependent Telecommunication Traffic Sponsor: Sprint Yong-Qing Lu David W. Petr Victor Frost Technical Report TISL-10230-4 Telecommunications and Information

More information

These outputs can be written in a more convenient form: with y(i) = Hc m (i) n(i) y(i) = (y(i); ; y K (i)) T ; c m (i) = (c m (i); ; c m K(i)) T and n

These outputs can be written in a more convenient form: with y(i) = Hc m (i) n(i) y(i) = (y(i); ; y K (i)) T ; c m (i) = (c m (i); ; c m K(i)) T and n Binary Codes for synchronous DS-CDMA Stefan Bruck, Ulrich Sorger Institute for Network- and Signal Theory Darmstadt University of Technology Merckstr. 25, 6428 Darmstadt, Germany Tel.: 49 65 629, Fax:

More information

Statistical Learning Notes III-Section 2.4.3

Statistical Learning Notes III-Section 2.4.3 Statistical Learning Notes III-Section 2.4.3 "Graphical" Spectral Features Stephen Vardeman Analytics Iowa LLC January 2019 Stephen Vardeman (Analytics Iowa LLC) Statistical Learning Notes III-Section

More information

P 1j. P jm. P ij. p m

P 1j. P jm. P ij. p m Analytic Model of Performance in Telecommunication Systems, Based on On-O Trac Sources with Self-Similar Behavior Lester Lipsky Department of Computer Science and Engineering University of Connecticut,

More information

Segmentation of Dynamic Scenes from the Multibody Fundamental Matrix

Segmentation of Dynamic Scenes from the Multibody Fundamental Matrix ECCV Workshop on Vision and Modeling of Dynamic Scenes, Copenhagen, Denmark, May 2002 Segmentation of Dynamic Scenes from the Multibody Fundamental Matrix René Vidal Dept of EECS, UC Berkeley Berkeley,

More information

Problem Set 9 Due: In class Tuesday, Nov. 27 Late papers will be accepted until 12:00 on Thursday (at the beginning of class).

Problem Set 9 Due: In class Tuesday, Nov. 27 Late papers will be accepted until 12:00 on Thursday (at the beginning of class). Math 3, Fall Jerry L. Kazdan Problem Set 9 Due In class Tuesday, Nov. 7 Late papers will be accepted until on Thursday (at the beginning of class).. Suppose that is an eigenvalue of an n n matrix A and

More information

Neural Controller. Plant. Plant. Critic. evaluation signal. Reinforcement Learning Controller

Neural Controller. Plant. Plant. Critic. evaluation signal. Reinforcement Learning Controller Neural Control Theory: an Overview J.A.K. Suykens, H. Bersini Katholieke Universiteit Leuven Department of Electrical Engineering, ESAT-SISTA Kardinaal Mercierlaan 94, B-3001 Leuven (Heverlee), Belgium

More information

Network Traffic Characteristic

Network Traffic Characteristic Network Traffic Characteristic Hojun Lee hlee02@purros.poly.edu 5/24/2002 EL938-Project 1 Outline Motivation What is self-similarity? Behavior of Ethernet traffic Behavior of WAN traffic Behavior of WWW

More information

The Singular Value Decomposition (SVD) and Principal Component Analysis (PCA)

The Singular Value Decomposition (SVD) and Principal Component Analysis (PCA) Chapter 5 The Singular Value Decomposition (SVD) and Principal Component Analysis (PCA) 5.1 Basics of SVD 5.1.1 Review of Key Concepts We review some key definitions and results about matrices that will

More information

STOCHASTIC MULTI-SCALE MODELLING OF SHORT- AND LONG-RANGE EFFECTS IN TEXTILE COMPOSITES BASED ON EXPERIMENTAL DATA

STOCHASTIC MULTI-SCALE MODELLING OF SHORT- AND LONG-RANGE EFFECTS IN TEXTILE COMPOSITES BASED ON EXPERIMENTAL DATA STOCHASTIC MULTI-SCALE MODELLING OF SHORT- AND LONG-RANGE EFFECTS IN TEXTILE COMPOSITES BASED ON EXPERIMENTAL DATA A. Vanaerschot 1, B. N. Cox 2, S. V. Lomov 3, D. Vandepitte 1 1 KU Leuven, Dept. of Mechanical

More information

Average Reward Parameters

Average Reward Parameters Simulation-Based Optimization of Markov Reward Processes: Implementation Issues Peter Marbach 2 John N. Tsitsiklis 3 Abstract We consider discrete time, nite state space Markov reward processes which depend

More information

THROUGHOUT the last few decades, the field of linear

THROUGHOUT the last few decades, the field of linear IEEE TRANSACTIONS ON AUTOMATIC CONTROL, VOL 50, NO 10, OCTOBER 2005 1509 Subspace Identification of Hammerstein Systems Using Least Squares Support Vector Machines Ivan Goethals, Kristiaan Pelckmans, Johan

More information

Contents. 6 Systems of First-Order Linear Dierential Equations. 6.1 General Theory of (First-Order) Linear Systems

Contents. 6 Systems of First-Order Linear Dierential Equations. 6.1 General Theory of (First-Order) Linear Systems Dierential Equations (part 3): Systems of First-Order Dierential Equations (by Evan Dummit, 26, v 2) Contents 6 Systems of First-Order Linear Dierential Equations 6 General Theory of (First-Order) Linear

More information

Generative Kernel PCA

Generative Kernel PCA ESA 8 proceedings, European Symposium on Artificial eural etworks, Computational Intelligence and Machine Learning. Bruges (Belgium), 5-7 April 8, i6doc.com publ., ISB 978-8758747-6. Generative Kernel

More information

1 Principal component analysis and dimensional reduction

1 Principal component analysis and dimensional reduction Linear Algebra Working Group :: Day 3 Note: All vector spaces will be finite-dimensional vector spaces over the field R. 1 Principal component analysis and dimensional reduction Definition 1.1. Given an

More information

The purpose of this paper is to demonstrate the exibility of the SPA

The purpose of this paper is to demonstrate the exibility of the SPA TIPP and the Spectral Expansion Method I. Mitrani 1, A. Ost 2, and M. Rettelbach 2 1 University of Newcastle upon Tyne 2 University of Erlangen-Nurnberg Summary. Stochastic Process Algebras (SPA) like

More information

arxiv: v2 [math.ds] 8 Dec 2014

arxiv: v2 [math.ds] 8 Dec 2014 arxiv:1406.4738v2 math.ds 8 Dec 2014 Some Special Cases in the Stability Analysis of Multi-Dimensional Time-Delay Systems Using The Matrix Lambert W function Rudy Cepeda-Gomez and Wim Michiels July 27,

More information

Markov Chains and Spectral Clustering

Markov Chains and Spectral Clustering Markov Chains and Spectral Clustering Ning Liu 1,2 and William J. Stewart 1,3 1 Department of Computer Science North Carolina State University, Raleigh, NC 27695-8206, USA. 2 nliu@ncsu.edu, 3 billy@ncsu.edu

More information

Submitted to IEEE Transactions on Computers, June Evaluating Dynamic Failure Probability for Streams with. (m; k)-firm Deadlines

Submitted to IEEE Transactions on Computers, June Evaluating Dynamic Failure Probability for Streams with. (m; k)-firm Deadlines Submitted to IEEE Transactions on Computers, June 1994 Evaluating Dynamic Failure Probability for Streams with (m; k)-firm Deadlines Moncef Hamdaoui and Parameswaran Ramanathan Department of Electrical

More information

Properties of Matrices and Operations on Matrices

Properties of Matrices and Operations on Matrices Properties of Matrices and Operations on Matrices A common data structure for statistical analysis is a rectangular array or matris. Rows represent individual observational units, or just observations,

More information

A Subspace Approach to Estimation of. Measurements 1. Carlos E. Davila. Electrical Engineering Department, Southern Methodist University

A Subspace Approach to Estimation of. Measurements 1. Carlos E. Davila. Electrical Engineering Department, Southern Methodist University EDICS category SP 1 A Subspace Approach to Estimation of Autoregressive Parameters From Noisy Measurements 1 Carlos E Davila Electrical Engineering Department, Southern Methodist University Dallas, Texas

More information

We use the overhead arrow to denote a column vector, i.e., a number with a direction. For example, in three-space, we write

We use the overhead arrow to denote a column vector, i.e., a number with a direction. For example, in three-space, we write 1 MATH FACTS 11 Vectors 111 Definition We use the overhead arrow to denote a column vector, ie, a number with a direction For example, in three-space, we write The elements of a vector have a graphical

More information