Using Matrix Analysis to Approach the Simulated Annealing Algorithm

Size: px
Start display at page:

Download "Using Matrix Analysis to Approach the Simulated Annealing Algorithm"

Transcription

1 Using Matrix Analysis to Approach the Simulated Annealing Algorithm Dennis I. Merino, Edgar N. Reyes Southeastern Louisiana University Hammond, LA 7040 and Carl Steidley Texas A&M University- Corpus Christi Corpus Christi, TX 784 An Overview of the Simulated Annealing Algorithm In [3, 4], the annealing of a solid is simulated and described by using the simulated annealing algorithm, a method based in combinatorial optimization. The subject of combinatorial optimization consists of a set of problems that are central to the disciplines of computer science and engineering. Annealing is the physical process of heating a solid and following it by a controlled cooling process. A simulated annealing of a dodecahedron, in particular, was analyzed in the earlier papers. It was pointed out in these papers that simulated annealing could be extended to other solids and plane figures such as polygons as well. Since the simulated annealing algorithm is stochastic in nature and has direct connections to Boltzman machines - a neural network, in this paper we shall generalize the results in the earlier papers and use matrix analysis techniques to simulate annealing based and modeled on a class of Boltzman machines which includes as a special case the simulated annealing of the dodecahedron. For clarity, we shall briefly review some aspects of Boltzman machines and the simulated annealing algorithm. Let (U, C) be a network consisting of units, U = {u,u,...,u n }, and a set of connections, C, consisting of unordered pairs {u i,u j }. A connection {u i,u j } Cis said to join u i to u j. Intrinsic to Boltzman machines are the notions of a connection strength s and a configuration k of the network (U, C); respectively, they are mappings s : C IR and k : U IR into the set of real numbers. A Boltzman machine is a network (U, C) with a given connection strength s. An objective of a Boltzman machine is to attain an overall measurement of desirability, see [, chapter 8]. Specifically, we shall seek to find an optimal configuration k opt in the space S of all configurations k that minimizes the consensus function C : S IR where C(k) = s({u, v})k(u)k(v). () {u,v} C To optimize (), we shall use the simulated annealing algorithm. There are several ways to implement this algorithm and each way depends on a neighborhood structure and a transition mechanism. The latter term suggests a link with Markov chains. A neighborhood structure is a mapping N : S P(S) where P(S) is the family of all subsets of S. A configuration l N(k) is called a neighbor of k. To optimize the consensus (), we shall need a mechanism which allows a configuration to change. Configurations may change from one trial to the next;

2 in the earlier papers, the configurations which simulate the temperatures on the faces of the dodecahedron change every minute. Given a configuration k, we shall stochastically generate a neighbor l N(k), with the neighborhood structure being defined at the outset, and then it will be determined whether l will replace k. Specifically, let X(m) be the configuration on the mth trial and let P k,l (m) =P (X(m) =l X(m ) = k) be the probability of accepting configuration l on the mth trial given that the configuration of the (m )th trial is k. Under certain conditions, such as those discussed in [, page 8, 4, or 46], the sequence of configurations generated by the simulated annealing algorithm asymptotically converges to an optimal configuration. The controlled cooling process shall be represented by a sequence {c m : m =0,,, 3,...} of real numbers. Following [3], the simulated annealing is described by the algorithm below Begin Simulated Annealing Algorithm Initialize: k; an intial configuration m = 0; a counter Do: generate l in N (k); if C(l) C(k), then k = l else if P k,l (m) > Random(0, ), then k = l m := m +; Until: Calculate Stop Criterion(c m ); end; Extending Simulated Annealing In [3], the set of units U d = {u i : i =,, 3,..., } represented the faces of a dodecahedron and the set of connections was C d = {{u i,u j } : i, j }. A configuration k of (U d, C d ) played the role of temperature. If at a particular minute the temperature on face u i is k(u i ), then on the second minute the temperatures on the faces are prescribed by N d (k) where N d (k)(u i )= 0 k(uj ), () the sum being taken over all units u j, except u i and the face opposite u i. (Since a dodecahedron is a regular solid, we have the notion of opposite faces.) Note, the neighborhood structure N d in [3] has special properties, for one, N d (k) P(S) is a singleton for each configuration k; in particular, N d (k) itself is a configuration. More importantly, since we can identify S with C, N d :C C is actually a linear operator. Also, the transition probability in [3] given by P k,l (m) = { if N (k) =l 0 otherwise (3) shows that a configuration k in one trial (or a minute) changes to the configuration N d (k) in the next trial (or minute); one can say that the transition mechanism is deterministic. In [3], it was

3 shown that when the connection strength s d is defined by s d ({u, v})= { if u = v if u v 6 (4) and k =(k,k,..., k ), then after several trials (or minutes) the mth configuration Nd m (k) shall ( be approximately k i,..., k i ), an optimal configuration minimizing (). Note, with a suitable basis we can represent the linear operator N d :C C by the matrix N d Let k =[k,k,...,k ] T be the column vector representing the temperatures of the faces of the dodecahedron on the first minute. Then, on the second minute the corresponding temperatures of the faces are the entries of the vector N d k. On the third minute, the temperatures are given by the vector N d (N d k)=nd (k). In general, on the (m +) th minute, the temperatures are given by Nd m k. A direct calculation reveals that lim N m m d =. (5) Hence, in each of the faces, the temperature will approach the average of the initial temperatures of all the twelve faces. 3

4 In this paper, we shall extend the results outlined in [3]. Let U = {u,u,...,u n } and let C = {{u i,u j } : i, j n} be a network representing a Boltzman machine, where n. For the connection strength s, we consider s({u, v})= { n n if u = v if u v. n (6) An optimal configuration k opt minimizing the consensus () with connection strength (6) necessarily and sufficiently minimizes n (k(u i ) µ(k)) (7) n over all configurations k where µ(k) = k(u n i ), see [3] for details regarding the case when n =. Next, we shall consider a neighborhood structure N whose values are not singletons but are infinite subsets of S. In particular, this implies that N is not a linear operator as was the case of the simulated annealing of the dodecahedron. Second, the transition probabilities P k,l (m) are stochastic and not deterministic as the one in (3). Let N =.. be the n-by-n matrix whose entries are all s. If necessary, we will emphasize the size of N by a subscript. For instance N is a -by- matrix. We say that an n-by-n matrix, Q, isdoubly stochastic if the entries of Q are nonnegative, and the sum of the entries in each row and column is. In the case of the simulated annealing of the dodecahedron, note that N d = (N 0 Q) where Q is the doubly stochastic matrix given by Q =. (8) In extending the results outlined in [3], we define the neighborhood structure according to { N (k) = n c (N n cq) : Q is a doubly stochastic matrix and 0 c< n }. (9) For that part of the program in section describing the simulated annealing algorithm, we need to define a mechanism of generating a neighbor l N(k) ofk. Let {E r : r =,, 3,..., n!} be the 4

5 set of all n-by-n permutation matrices. Note, by a Theorem due to Birkhoff [, Theorem 8.7.], each doubly stochastic matrix Q can be expressed in the form n! where n! Q = α r E r (0) α r =,α r 0. This implies that given c< n, we can stochastically generate a doubly stochastic matrix Q by stochastically choosing an n!-tuple (α r ) and then constructing the matrix in (0). Furthermore, to implement the simulated annealing algorithm we shall have to define a transition probability. For one, let P k,l (m) =α () where l N(k) is given by l = ( n! N n c α r E r ). n c Suppose k is an initial configuration of a Boltzman machine (U, C). By realizing k as a column vector, then the configuration in the second trial is n c (N n c Q )k and the configuration in the third trial is ( n c (N n c Q ) )( n c (N n c Q ) ) k for some 0 c,c < n and doubly stochastic matrices Q,Q. In this paper, provided each c i satisfes 0 c i < n ɛ for some ɛ>0, we shall show that after several trials the m th configuration shall be approximately N n n(k), i.e., m lim m n c i (N n c i Q i )(k) = n N n(k). () As a special case, when n =, c i =, and each Q i is the matrix in (8) for all i, we obtain the case of the simulated annealing of the dodecahedron in [3], that is, m n c i (N n c i Q i )(k) = Nd m (k). In Appendix A we shall verify () and in Appendix B we provide an example using the software package GAP to simulate the annealing of a dodecahedron [5]. GAP (acronym for Groups, Algorithms, and Programming) is a software package suitable for use by undergraduates in advanced classes such as abstract algebra, linear algebra, and number theory. 3 Summary This paper generalizes the simulated annealing process of the dodecahedron in [3, 4]. We have chosen U = {u,u,...,u n } as the set of units and C = {{u i,u j } : i, j n} as the set of connections of a Boltzman machine. We assume that the connection strength of this neural network (U, C) is the one given by (6). Moreover, we have chosen the neighborhood structure N : S P(S) defined in (9). Given an initial configuration k, we generate a neighbor l N(k) by n! choosing stochastically or randomly an n!-tuple (α r ) of non-negative scalars satisfying α r =, α r 0 for all r and then by setting l to be the matrix in the right-hand side of (0). The transition probability P k,l we chose is the one in (). The choice for P k,l (m) is not crucial at this juncture since () shall hold for any other transition probability but we decided to have chosen one for the sake of completeness. 5

6 To simulate a controlled cooling process, we shall choose a sequence {c m : m =,, 3...} of real numbers satisfying lim c m = 0. The choice of this limit, namely 0, is motivated by the simulated m annealing of the dodecahedron. We quote from [3] some features of the simulated annealing of the dodecahedron: Initially, for each i {,..., } the face u i is labeled with the number k(u i ) and on the second minute the number on face u i becomes the average of the first minute s numbers except those that were on u i and on the face opposite u i, i.e. on the second minute the number on face u i is k(u 0 j ) where the sum is taken over all faces u j except the one on face u i and on the face opposite u i. j We see that when c =, the temperature on a face in the next trial or minute is given by the average of 0 temperatures of the current trial or minute. If we prescribe the temperature on a face in the next trial to be the average of the temperatures of the five neighboring faces (a face in a dodecahedron has five surrounding faces) in the current trial, then the value of c is 7. This suggests to us that a slow cooling process is simulated when the averaging of the temperatures begins with a few faces and then gradually increased to more faces. Thus, we assume that the sequence {c m : m =,, 3,...} in section describing the simulated annealing algorithm is decreasing and converges to 0. { In the general case, we shall show that the sequence of configurations, namely, k, n c (N n c Q )k, ( n c 3 (N n c 3 Q 3 ) )( n c (N n c Q ) ) k, } converges to N n n provided 0 c i n ɛ for some ɛ. Thus, if the initial configuration is k then after several trials the limiting configuration has a constant value at each unit, the value being the average of {k(u i ): i =,, 3,..., n}. For the simulated annealing, this implies that the temperatures evenly distributes over all the units. Note, we do not have any discussion about the rate of convergence which may be dependent on the choice of the transition probability. 4 Appendix A In this section we shall present a short proof supporting our claim (). Theorem Let n be an integer. Suppose 0 c i < n ɛ and Q i is a doubly stochastic matrix m for all i where ɛ is some positive real number. Then lim (N c i Q i )= m n c i n N. Proof. Notice that N = nn, and in general N l = n l N for each positive integer l. A direct computation reveals that NQ = QN = N for any n-by-n doubly stochastic matrix Q. Note, the product of two doubly stochastic matrices is again doubly stochastic; therefore, the entries of m m Q i are bounded. Rewriting the product (N c i Q i ), we obtain n c i = m n c i [(N c Q ) (N c m Q m )] 6

7 = m m N m c i N (m ) + c i c i N (m ) c i c i n c i i <i i <i <i 3

8 connections. In carrying out the calculations involved in the simulated annealing algorithm, we will use a software for computational discrete algebra, namely, GAP [5]. Given a configuration k of the network (U d, C d ), the set N (k) of neighbors of k is the set defined in (9). Let us consider the permutation matrices {E r : r =,..., 6} generated by the permutations (4, 8, ) and (8, ). This is a small subset of all -by- permutation matrices. This subset is actually a dihedral group. To stochastically generate a neighbor l N(k), see algorithm before the beginning of section and (0), we will randomly select a stochastic -by- matrix of the form 6 6 Q = α r E r where α r =,α r = i r 3 0, i r {0,,, 3}. Moreover, if l = n c (N n cq) then the transition probability shall be chosen according to (), that is, P k,l = α. If the initial configuration is k =[,, 3, 4, 5, 6, 7, 8, 9, 0,, ] and the sequence c i representing temperature which is slowly decreasing is given by c i = where i =,..., 0, then the i [ final output of the algorithm described just before section is approximately the configuration 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, ] Note, the average of the numbers,.., is 6.5 which is approximately 37. This supports our claim, as stated in the theorem in appendix A, 50 that the sequence of configurations generated by the simulated annealing converges to an optimal configuration minimizing (7). The following are the commands we keyed in when we were using GAP. gap > g := Group((4, 8, ), (8, )); ; gap > gl := Elements(g); ; gap > Size(gl); 6 gap > L := [ ]; ; gap > for i in [0..3] do for j in [..3]do gap > if i <= j then Append(L, [i/j]); gap > fi; od; od gap > S := Set(L); ; T := T uples(s, 6); ; gap > c := [ ]; ; gap > for i in [..Length(T )] do gap > z := List([..6],j >T[i][j]); gap > if Sum(z) =then Append(c, [z]) ; fi; od; gap > cons := function(x) gap > local ave; gap > ave := Sum(List([..],j >x[j]))/; gap > return Sum(List([..],j > (x[j] ave) )); end; gap > r := [ ]; ; q := [ ]; ; l := [ ]; ; k := [,, 3, 4, 5, 6, 7, 8, 9, 0,, ]; ; gap > for i in [..0] do gap > j := i ; 8

9 References gap > coef := Random(c); gap > for m in [..] do gap > for p in [..6] do gap > r[p] :=coef[p] k[m (gl[p] ( ))]; od gap > l[m]:=sum(r); r := [ ]; gap > q[m] := (/( j)) (Sum(k) j l[m]); od; gap > if cons(q) <= cons(k) then k := q; gap > elif coef[] > AbsInt(Random([..]))/ then k := q; gap > fi; od; gap > final := [ ]; ; gap > for i in [..] do gap > final[i]:= Int(00 k[i])/00; od; gap > final; [37 ] 50 [] Aarts, E. and Korst, J., Simulated Annealing and Boltzman Machines, Wiley-Interscience Series in Discrete Mathematics and Optimization, John Wiley and Sons, 989. [] Horn, R.A. and Johnson, C.R., Matrix Analysis, Cambridge University Press, 990. [3] Reyes, E.N. and Steidley, C., A GAP Approach to the Simulated Annealing Algorithm, Computers in Education Journal of the American Society for Engineering Education, Vol. 7, No.3 (997), [4] Reyes, E.N. and Steidley, C., A Theoretical Discussion of the Mathematics Underlying A GAP Approach to the Simulated Annealing Algorithm, to appear in the Computers in Education Journal of the American Society for Engineering Education. [5] M. Schönert et al., GAP-Groups, Algorithms, and Programming, Version 3 Release 4, Lehrstuhl D für Mathematik, RWTH, Aachen, Germany,

Alternative Characterization of Ergodicity for Doubly Stochastic Chains

Alternative Characterization of Ergodicity for Doubly Stochastic Chains Alternative Characterization of Ergodicity for Doubly Stochastic Chains Behrouz Touri and Angelia Nedić Abstract In this paper we discuss the ergodicity of stochastic and doubly stochastic chains. We define

More information

Random Walks on Graphs. One Concrete Example of a random walk Motivation applications

Random Walks on Graphs. One Concrete Example of a random walk Motivation applications Random Walks on Graphs Outline One Concrete Example of a random walk Motivation applications shuffling cards universal traverse sequence self stabilizing token management scheme random sampling enumeration

More information

Simulated Annealing. Local Search. Cost function. Solution space

Simulated Annealing. Local Search. Cost function. Solution space Simulated Annealing Hill climbing Simulated Annealing Local Search Cost function? Solution space Annealing Annealing is a thermal process for obtaining low energy states of a solid in a heat bath. The

More information

Lecture 20 : Markov Chains

Lecture 20 : Markov Chains CSCI 3560 Probability and Computing Instructor: Bogdan Chlebus Lecture 0 : Markov Chains We consider stochastic processes. A process represents a system that evolves through incremental changes called

More information

RATIONAL REALIZATION OF MAXIMUM EIGENVALUE MULTIPLICITY OF SYMMETRIC TREE SIGN PATTERNS. February 6, 2006

RATIONAL REALIZATION OF MAXIMUM EIGENVALUE MULTIPLICITY OF SYMMETRIC TREE SIGN PATTERNS. February 6, 2006 RATIONAL REALIZATION OF MAXIMUM EIGENVALUE MULTIPLICITY OF SYMMETRIC TREE SIGN PATTERNS ATOSHI CHOWDHURY, LESLIE HOGBEN, JUDE MELANCON, AND RANA MIKKELSON February 6, 006 Abstract. A sign pattern is a

More information

Boolean Inner-Product Spaces and Boolean Matrices

Boolean Inner-Product Spaces and Boolean Matrices Boolean Inner-Product Spaces and Boolean Matrices Stan Gudder Department of Mathematics, University of Denver, Denver CO 80208 Frédéric Latrémolière Department of Mathematics, University of Denver, Denver

More information

Markov Chains. As part of Interdisciplinary Mathematical Modeling, By Warren Weckesser Copyright c 2006.

Markov Chains. As part of Interdisciplinary Mathematical Modeling, By Warren Weckesser Copyright c 2006. Markov Chains As part of Interdisciplinary Mathematical Modeling, By Warren Weckesser Copyright c 2006 1 Introduction A (finite) Markov chain is a process with a finite number of states (or outcomes, or

More information

Intrinsic products and factorizations of matrices

Intrinsic products and factorizations of matrices Available online at www.sciencedirect.com Linear Algebra and its Applications 428 (2008) 5 3 www.elsevier.com/locate/laa Intrinsic products and factorizations of matrices Miroslav Fiedler Academy of Sciences

More information

Single Solution-based Metaheuristics

Single Solution-based Metaheuristics Parallel Cooperative Optimization Research Group Single Solution-based Metaheuristics E-G. Talbi Laboratoire d Informatique Fondamentale de Lille Single solution-based metaheuristics Improvement of a solution.

More information

On asymptotic behavior of a finite Markov chain

On asymptotic behavior of a finite Markov chain 1 On asymptotic behavior of a finite Markov chain Alina Nicolae Department of Mathematical Analysis Probability. University Transilvania of Braşov. Romania. Keywords: convergence, weak ergodicity, strong

More information

arxiv: v1 [math.co] 10 Aug 2016

arxiv: v1 [math.co] 10 Aug 2016 POLYTOPES OF STOCHASTIC TENSORS HAIXIA CHANG 1, VEHBI E. PAKSOY 2 AND FUZHEN ZHANG 2 arxiv:1608.03203v1 [math.co] 10 Aug 2016 Abstract. Considering n n n stochastic tensors (a ijk ) (i.e., nonnegative

More information

Multi-coloring and Mycielski s construction

Multi-coloring and Mycielski s construction Multi-coloring and Mycielski s construction Tim Meagher Fall 2010 Abstract We consider a number of related results taken from two papers one by W. Lin [1], and the other D. C. Fisher[2]. These articles

More information

MATH 56A: STOCHASTIC PROCESSES CHAPTER 1

MATH 56A: STOCHASTIC PROCESSES CHAPTER 1 MATH 56A: STOCHASTIC PROCESSES CHAPTER. Finite Markov chains For the sake of completeness of these notes I decided to write a summary of the basic concepts of finite Markov chains. The topics in this chapter

More information

642:550, Summer 2004, Supplement 6 The Perron-Frobenius Theorem. Summer 2004

642:550, Summer 2004, Supplement 6 The Perron-Frobenius Theorem. Summer 2004 642:550, Summer 2004, Supplement 6 The Perron-Frobenius Theorem. Summer 2004 Introduction Square matrices whose entries are all nonnegative have special properties. This was mentioned briefly in Section

More information

Markov Chains, Stochastic Processes, and Matrix Decompositions

Markov Chains, Stochastic Processes, and Matrix Decompositions Markov Chains, Stochastic Processes, and Matrix Decompositions 5 May 2014 Outline 1 Markov Chains Outline 1 Markov Chains 2 Introduction Perron-Frobenius Matrix Decompositions and Markov Chains Spectral

More information

Definition A finite Markov chain is a memoryless homogeneous discrete stochastic process with a finite number of states.

Definition A finite Markov chain is a memoryless homogeneous discrete stochastic process with a finite number of states. Chapter 8 Finite Markov Chains A discrete system is characterized by a set V of states and transitions between the states. V is referred to as the state space. We think of the transitions as occurring

More information

0 Sets and Induction. Sets

0 Sets and Induction. Sets 0 Sets and Induction Sets A set is an unordered collection of objects, called elements or members of the set. A set is said to contain its elements. We write a A to denote that a is an element of the set

More information

Computational statistics

Computational statistics Computational statistics Combinatorial optimization Thierry Denœux February 2017 Thierry Denœux Computational statistics February 2017 1 / 37 Combinatorial optimization Assume we seek the maximum of f

More information

THE N-VALUE GAME OVER Z AND R

THE N-VALUE GAME OVER Z AND R THE N-VALUE GAME OVER Z AND R YIDA GAO, MATT REDMOND, ZACH STEWARD Abstract. The n-value game is an easily described mathematical diversion with deep underpinnings in dynamical systems analysis. We examine

More information

Markov decision processes and interval Markov chains: exploiting the connection

Markov decision processes and interval Markov chains: exploiting the connection Markov decision processes and interval Markov chains: exploiting the connection Mingmei Teo Supervisors: Prof. Nigel Bean, Dr Joshua Ross University of Adelaide July 10, 2013 Intervals and interval arithmetic

More information

Finite-Horizon Statistics for Markov chains

Finite-Horizon Statistics for Markov chains Analyzing FSDT Markov chains Friday, September 30, 2011 2:03 PM Simulating FSDT Markov chains, as we have said is very straightforward, either by using probability transition matrix or stochastic update

More information

CHAPTER 6. Markov Chains

CHAPTER 6. Markov Chains CHAPTER 6 Markov Chains 6.1. Introduction A(finite)Markovchainisaprocess withafinitenumberofstates (or outcomes, or events) in which the probability of being in a particular state at step n+1depends only

More information

3. Vector spaces 3.1 Linear dependence and independence 3.2 Basis and dimension. 5. Extreme points and basic feasible solutions

3. Vector spaces 3.1 Linear dependence and independence 3.2 Basis and dimension. 5. Extreme points and basic feasible solutions A. LINEAR ALGEBRA. CONVEX SETS 1. Matrices and vectors 1.1 Matrix operations 1.2 The rank of a matrix 2. Systems of linear equations 2.1 Basic solutions 3. Vector spaces 3.1 Linear dependence and independence

More information

6 Markov Chain Monte Carlo (MCMC)

6 Markov Chain Monte Carlo (MCMC) 6 Markov Chain Monte Carlo (MCMC) The underlying idea in MCMC is to replace the iid samples of basic MC methods, with dependent samples from an ergodic Markov chain, whose limiting (stationary) distribution

More information

LECTURE 10: REVIEW OF POWER SERIES. 1. Motivation

LECTURE 10: REVIEW OF POWER SERIES. 1. Motivation LECTURE 10: REVIEW OF POWER SERIES By definition, a power series centered at x 0 is a series of the form where a 0, a 1,... and x 0 are constants. For convenience, we shall mostly be concerned with the

More information

Distributed Randomized Algorithms for the PageRank Computation Hideaki Ishii, Member, IEEE, and Roberto Tempo, Fellow, IEEE

Distributed Randomized Algorithms for the PageRank Computation Hideaki Ishii, Member, IEEE, and Roberto Tempo, Fellow, IEEE IEEE TRANSACTIONS ON AUTOMATIC CONTROL, VOL. 55, NO. 9, SEPTEMBER 2010 1987 Distributed Randomized Algorithms for the PageRank Computation Hideaki Ishii, Member, IEEE, and Roberto Tempo, Fellow, IEEE Abstract

More information

Aunu Integer Sequence as Non-Associative Structure and Their Graph Theoretic Properties

Aunu Integer Sequence as Non-Associative Structure and Their Graph Theoretic Properties Advances in Pure Mathematics, 2016, 6, 409-419 Published Online May 2016 in SciRes. http://www.scirp.org/journal/apm http://dx.doi.org/10.4236/apm.2016.66028 Aunu Integer Sequence as Non-Associative Structure

More information

Markov Chains. Andreas Klappenecker by Andreas Klappenecker. All rights reserved. Texas A&M University

Markov Chains. Andreas Klappenecker by Andreas Klappenecker. All rights reserved. Texas A&M University Markov Chains Andreas Klappenecker Texas A&M University 208 by Andreas Klappenecker. All rights reserved. / 58 Stochastic Processes A stochastic process X tx ptq: t P T u is a collection of random variables.

More information

M17 MAT25-21 HOMEWORK 6

M17 MAT25-21 HOMEWORK 6 M17 MAT25-21 HOMEWORK 6 DUE 10:00AM WEDNESDAY SEPTEMBER 13TH 1. To Hand In Double Series. The exercises in this section will guide you to complete the proof of the following theorem: Theorem 1: Absolute

More information

P(X 0 = j 0,... X nk = j k )

P(X 0 = j 0,... X nk = j k ) Introduction to Probability Example Sheet 3 - Michaelmas 2006 Michael Tehranchi Problem. Let (X n ) n 0 be a homogeneous Markov chain on S with transition matrix P. Given a k N, let Z n = X kn. Prove that

More information

RANK AND PERIMETER PRESERVER OF RANK-1 MATRICES OVER MAX ALGEBRA

RANK AND PERIMETER PRESERVER OF RANK-1 MATRICES OVER MAX ALGEBRA Discussiones Mathematicae General Algebra and Applications 23 (2003 ) 125 137 RANK AND PERIMETER PRESERVER OF RANK-1 MATRICES OVER MAX ALGEBRA Seok-Zun Song and Kyung-Tae Kang Department of Mathematics,

More information

Chapter 7 Network Flow Problems, I

Chapter 7 Network Flow Problems, I Chapter 7 Network Flow Problems, I Network flow problems are the most frequently solved linear programming problems. They include as special cases, the assignment, transportation, maximum flow, and shortest

More information

http://www.math.uah.edu/stat/markov/.xhtml 1 of 9 7/16/2009 7:20 AM Virtual Laboratories > 16. Markov Chains > 1 2 3 4 5 6 7 8 9 10 11 12 1. A Markov process is a random process in which the future is

More information

Kernels of Directed Graph Laplacians. J. S. Caughman and J.J.P. Veerman

Kernels of Directed Graph Laplacians. J. S. Caughman and J.J.P. Veerman Kernels of Directed Graph Laplacians J. S. Caughman and J.J.P. Veerman Department of Mathematics and Statistics Portland State University PO Box 751, Portland, OR 97207. caughman@pdx.edu, veerman@pdx.edu

More information

Vertex colorings of graphs without short odd cycles

Vertex colorings of graphs without short odd cycles Vertex colorings of graphs without short odd cycles Andrzej Dudek and Reshma Ramadurai Department of Mathematical Sciences Carnegie Mellon University Pittsburgh, PA 1513, USA {adudek,rramadur}@andrew.cmu.edu

More information

The concentration of the chromatic number of random graphs

The concentration of the chromatic number of random graphs The concentration of the chromatic number of random graphs Noga Alon Michael Krivelevich Abstract We prove that for every constant δ > 0 the chromatic number of the random graph G(n, p) with p = n 1/2

More information

On Expected Gaussian Random Determinants

On Expected Gaussian Random Determinants On Expected Gaussian Random Determinants Moo K. Chung 1 Department of Statistics University of Wisconsin-Madison 1210 West Dayton St. Madison, WI 53706 Abstract The expectation of random determinants whose

More information

Modeling and Stability Analysis of a Communication Network System

Modeling and Stability Analysis of a Communication Network System Modeling and Stability Analysis of a Communication Network System Zvi Retchkiman Königsberg Instituto Politecnico Nacional e-mail: mzvi@cic.ipn.mx Abstract In this work, the modeling and stability problem

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

RESEARCH ARTICLE. An extension of the polytope of doubly stochastic matrices

RESEARCH ARTICLE. An extension of the polytope of doubly stochastic matrices Linear and Multilinear Algebra Vol. 00, No. 00, Month 200x, 1 15 RESEARCH ARTICLE An extension of the polytope of doubly stochastic matrices Richard A. Brualdi a and Geir Dahl b a Department of Mathematics,

More information

Row and Column Distributions of Letter Matrices

Row and Column Distributions of Letter Matrices College of William and Mary W&M ScholarWorks Undergraduate Honors Theses Theses, Dissertations, & Master Projects 5-2016 Row and Column Distributions of Letter Matrices Xiaonan Hu College of William and

More information

A hierarchical network formation model

A hierarchical network formation model Available online at www.sciencedirect.com Electronic Notes in Discrete Mathematics 50 (2015) 379 384 www.elsevier.com/locate/endm A hierarchical network formation model Omid Atabati a,1 Babak Farzad b,2

More information

Theoretical Computer Science

Theoretical Computer Science Theoretical Computer Science 406 008) 3 4 Contents lists available at ScienceDirect Theoretical Computer Science journal homepage: www.elsevier.com/locate/tcs Discrete sets with minimal moment of inertia

More information

Reaching a Consensus in a Dynamically Changing Environment - Convergence Rates, Measurement Delays and Asynchronous Events

Reaching a Consensus in a Dynamically Changing Environment - Convergence Rates, Measurement Delays and Asynchronous Events Reaching a Consensus in a Dynamically Changing Environment - Convergence Rates, Measurement Delays and Asynchronous Events M. Cao Yale Univesity A. S. Morse Yale University B. D. O. Anderson Australia

More information

Quivers of Period 2. Mariya Sardarli Max Wimberley Heyi Zhu. November 26, 2014

Quivers of Period 2. Mariya Sardarli Max Wimberley Heyi Zhu. November 26, 2014 Quivers of Period 2 Mariya Sardarli Max Wimberley Heyi Zhu ovember 26, 2014 Abstract A quiver with vertices labeled from 1,..., n is said to have period 2 if the quiver obtained by mutating at 1 and then

More information

POSITIVE PARTIAL REALIZATION PROBLEM FOR LINEAR DISCRETE TIME SYSTEMS

POSITIVE PARTIAL REALIZATION PROBLEM FOR LINEAR DISCRETE TIME SYSTEMS Int J Appl Math Comput Sci 7 Vol 7 No 65 7 DOI: 478/v6-7-5- POSITIVE PARTIAL REALIZATION PROBLEM FOR LINEAR DISCRETE TIME SYSTEMS TADEUSZ KACZOREK Faculty of Electrical Engineering Białystok Technical

More information

MATRIX GENERATORS FOR THE REE GROUPS 2 G 2 (q)

MATRIX GENERATORS FOR THE REE GROUPS 2 G 2 (q) MATRIX GENERATORS FOR THE REE GROUPS 2 G 2 (q) Gregor Kemper Frank Lübeck and Kay Magaard May 18 2000 For the purposes of [K] and [KM] it became necessary to have 7 7 matrix generators for a Sylow-3-subgroup

More information

b jσ(j), Keywords: Decomposable numerical range, principal character AMS Subject Classification: 15A60

b jσ(j), Keywords: Decomposable numerical range, principal character AMS Subject Classification: 15A60 On the Hu-Hurley-Tam Conjecture Concerning The Generalized Numerical Range Che-Man Cheng Faculty of Science and Technology, University of Macau, Macau. E-mail: fstcmc@umac.mo and Chi-Kwong Li Department

More information

Reinforcement Learning

Reinforcement Learning Reinforcement Learning March May, 2013 Schedule Update Introduction 03/13/2015 (10:15-12:15) Sala conferenze MDPs 03/18/2015 (10:15-12:15) Sala conferenze Solving MDPs 03/20/2015 (10:15-12:15) Aula Alpha

More information

Linear equations in linear algebra

Linear equations in linear algebra Linear equations in linear algebra Samy Tindel Purdue University Differential equations and linear algebra - MA 262 Taken from Differential equations and linear algebra Pearson Collections Samy T. Linear

More information

A New Approximation Algorithm for the Asymmetric TSP with Triangle Inequality By Markus Bläser

A New Approximation Algorithm for the Asymmetric TSP with Triangle Inequality By Markus Bläser A New Approximation Algorithm for the Asymmetric TSP with Triangle Inequality By Markus Bläser Presented By: Chris Standish chriss@cs.tamu.edu 23 November 2005 1 Outline Problem Definition Frieze s Generic

More information

Some Results Concerning Uniqueness of Triangle Sequences

Some Results Concerning Uniqueness of Triangle Sequences Some Results Concerning Uniqueness of Triangle Sequences T. Cheslack-Postava A. Diesl M. Lepinski A. Schuyler August 12 1999 Abstract In this paper we will begin by reviewing the triangle iteration. We

More information

Selecting Efficient Correlated Equilibria Through Distributed Learning. Jason R. Marden

Selecting Efficient Correlated Equilibria Through Distributed Learning. Jason R. Marden 1 Selecting Efficient Correlated Equilibria Through Distributed Learning Jason R. Marden Abstract A learning rule is completely uncoupled if each player s behavior is conditioned only on his own realized

More information

What is a vector in hyperbolic geometry? And, what is a hyperbolic linear transformation?

What is a vector in hyperbolic geometry? And, what is a hyperbolic linear transformation? 0 What is a vector in hyperbolic geometry? And, what is a hyperbolic linear transformation? Ken Li, Dennis Merino, and Edgar N. Reyes Southeastern Louisiana University Hammond, LA USA 70402 1 Introduction

More information

Increments of Random Partitions

Increments of Random Partitions Increments of Random Partitions Şerban Nacu January 2, 2004 Abstract For any partition of {1, 2,...,n} we define its increments X i, 1 i n by X i =1ifi is the smallest element in the partition block that

More information

Orthogonal Arrays & Codes

Orthogonal Arrays & Codes Orthogonal Arrays & Codes Orthogonal Arrays - Redux An orthogonal array of strength t, a t-(v,k,λ)-oa, is a λv t x k array of v symbols, such that in any t columns of the array every one of the possible

More information

On the number of cycles in a graph with restricted cycle lengths

On the number of cycles in a graph with restricted cycle lengths On the number of cycles in a graph with restricted cycle lengths Dániel Gerbner, Balázs Keszegh, Cory Palmer, Balázs Patkós arxiv:1610.03476v1 [math.co] 11 Oct 2016 October 12, 2016 Abstract Let L be a

More information

INVARIANTS OF TWIST-WISE FLOW EQUIVALENCE

INVARIANTS OF TWIST-WISE FLOW EQUIVALENCE ELECTRONIC RESEARCH ANNOUNCEMENTS OF THE AMERICAN MATHEMATICAL SOCIETY Volume 3, Pages 126 130 (December 17, 1997) S 1079-6762(97)00037-1 INVARIANTS OF TWIST-WISE FLOW EQUIVALENCE MICHAEL C. SULLIVAN (Communicated

More information

Lecture 2: From Classical to Quantum Model of Computation

Lecture 2: From Classical to Quantum Model of Computation CS 880: Quantum Information Processing 9/7/10 Lecture : From Classical to Quantum Model of Computation Instructor: Dieter van Melkebeek Scribe: Tyson Williams Last class we introduced two models for deterministic

More information

CS 781 Lecture 9 March 10, 2011 Topics: Local Search and Optimization Metropolis Algorithm Greedy Optimization Hopfield Networks Max Cut Problem Nash

CS 781 Lecture 9 March 10, 2011 Topics: Local Search and Optimization Metropolis Algorithm Greedy Optimization Hopfield Networks Max Cut Problem Nash CS 781 Lecture 9 March 10, 2011 Topics: Local Search and Optimization Metropolis Algorithm Greedy Optimization Hopfield Networks Max Cut Problem Nash Equilibrium Price of Stability Coping With NP-Hardness

More information

21-301, Spring 2019 Homework 4 Solutions

21-301, Spring 2019 Homework 4 Solutions 21-301, Spring 2019 Homework 4 Solutions Michael Anastos Not to be published without written consent of the author or the instructors of the course. (Please contact me if you find any errors!) Problem

More information

THE DYNAMICS OF SUCCESSIVE DIFFERENCES OVER Z AND R

THE DYNAMICS OF SUCCESSIVE DIFFERENCES OVER Z AND R THE DYNAMICS OF SUCCESSIVE DIFFERENCES OVER Z AND R YIDA GAO, MATT REDMOND, ZACH STEWARD Abstract. The n-value game is a dynamical system defined by a method of iterated differences. In this paper, we

More information

5. Simulated Annealing 5.1 Basic Concepts. Fall 2010 Instructor: Dr. Masoud Yaghini

5. Simulated Annealing 5.1 Basic Concepts. Fall 2010 Instructor: Dr. Masoud Yaghini 5. Simulated Annealing 5.1 Basic Concepts Fall 2010 Instructor: Dr. Masoud Yaghini Outline Introduction Real Annealing and Simulated Annealing Metropolis Algorithm Template of SA A Simple Example References

More information

RESEARCH STATEMENT-ERIC SAMANSKY

RESEARCH STATEMENT-ERIC SAMANSKY RESEARCH STATEMENT-ERIC SAMANSKY Introduction The main topic of my work is geometric measure theory. Specifically, I look at the convergence of certain probability measures, called Gibbs measures, on fractals

More information

Solving Homogeneous Systems with Sub-matrices

Solving Homogeneous Systems with Sub-matrices Pure Mathematical Sciences, Vol 7, 218, no 1, 11-18 HIKARI Ltd, wwwm-hikaricom https://doiorg/112988/pms218843 Solving Homogeneous Systems with Sub-matrices Massoud Malek Mathematics, California State

More information

The spectra of super line multigraphs

The spectra of super line multigraphs The spectra of super line multigraphs Jay Bagga Department of Computer Science Ball State University Muncie, IN jbagga@bsuedu Robert B Ellis Department of Applied Mathematics Illinois Institute of Technology

More information

Reaching a Consensus in a Dynamically Changing Environment A Graphical Approach

Reaching a Consensus in a Dynamically Changing Environment A Graphical Approach Reaching a Consensus in a Dynamically Changing Environment A Graphical Approach M. Cao Yale Univesity A. S. Morse Yale University B. D. O. Anderson Australia National University and National ICT Australia

More information

arxiv: v1 [math.pr] 21 Mar 2014

arxiv: v1 [math.pr] 21 Mar 2014 Asymptotic distribution of two-protected nodes in ternary search trees Cecilia Holmgren Svante Janson March 2, 24 arxiv:4.557v [math.pr] 2 Mar 24 Abstract We study protected nodes in m-ary search trees,

More information

Analytic Number Theory Solutions

Analytic Number Theory Solutions Analytic Number Theory Solutions Sean Li Cornell University sxl6@cornell.edu Jan. 03 Introduction This document is a work-in-progress solution manual for Tom Apostol s Introduction to Analytic Number Theory.

More information

Signal Recovery from Permuted Observations

Signal Recovery from Permuted Observations EE381V Course Project Signal Recovery from Permuted Observations 1 Problem Shanshan Wu (sw33323) May 8th, 2015 We start with the following problem: let s R n be an unknown n-dimensional real-valued signal,

More information

Higher Spin Alternating Sign Matrices

Higher Spin Alternating Sign Matrices Higher Spin Alternating Sign Matrices Roger E. Behrend and Vincent A. Knight School of Mathematics, Cardiff University, Cardiff, CF24 4AG, UK behrendr@cardiff.ac.uk, knightva@cardiff.ac.uk Submitted: Aug

More information

ORDERS OF ELEMENTS IN A GROUP

ORDERS OF ELEMENTS IN A GROUP ORDERS OF ELEMENTS IN A GROUP KEITH CONRAD 1. Introduction Let G be a group and g G. We say g has finite order if g n = e for some positive integer n. For example, 1 and i have finite order in C, since

More information

Equivalence of Regular Expressions and FSMs

Equivalence of Regular Expressions and FSMs Equivalence of Regular Expressions and FSMs Greg Plaxton Theory in Programming Practice, Spring 2005 Department of Computer Science University of Texas at Austin Regular Language Recall that a language

More information

Chapter 1: Systems of linear equations and matrices. Section 1.1: Introduction to systems of linear equations

Chapter 1: Systems of linear equations and matrices. Section 1.1: Introduction to systems of linear equations Chapter 1: Systems of linear equations and matrices Section 1.1: Introduction to systems of linear equations Definition: A linear equation in n variables can be expressed in the form a 1 x 1 + a 2 x 2

More information

Relative Improvement by Alternative Solutions for Classes of Simple Shortest Path Problems with Uncertain Data

Relative Improvement by Alternative Solutions for Classes of Simple Shortest Path Problems with Uncertain Data Relative Improvement by Alternative Solutions for Classes of Simple Shortest Path Problems with Uncertain Data Part II: Strings of Pearls G n,r with Biased Perturbations Jörg Sameith Graduiertenkolleg

More information

6. APPLICATION TO THE TRAVELING SALESMAN PROBLEM

6. APPLICATION TO THE TRAVELING SALESMAN PROBLEM 6. Application to the Traveling Salesman Problem 92 6. APPLICATION TO THE TRAVELING SALESMAN PROBLEM The properties that have the most significant influence on the maps constructed by Kohonen s algorithm

More information

(x 1 +x 2 )(x 1 x 2 )+(x 2 +x 3 )(x 2 x 3 )+(x 3 +x 1 )(x 3 x 1 ).

(x 1 +x 2 )(x 1 x 2 )+(x 2 +x 3 )(x 2 x 3 )+(x 3 +x 1 )(x 3 x 1 ). CMPSCI611: Verifying Polynomial Identities Lecture 13 Here is a problem that has a polynomial-time randomized solution, but so far no poly-time deterministic solution. Let F be any field and let Q(x 1,...,

More information

THE REAL NUMBERS Chapter #4

THE REAL NUMBERS Chapter #4 FOUNDATIONS OF ANALYSIS FALL 2008 TRUE/FALSE QUESTIONS THE REAL NUMBERS Chapter #4 (1) Every element in a field has a multiplicative inverse. (2) In a field the additive inverse of 1 is 0. (3) In a field

More information

Strictly Positive Definite Functions on a Real Inner Product Space

Strictly Positive Definite Functions on a Real Inner Product Space Strictly Positive Definite Functions on a Real Inner Product Space Allan Pinkus Abstract. If ft) = a kt k converges for all t IR with all coefficients a k 0, then the function f< x, y >) is positive definite

More information

Eigenvalues, random walks and Ramanujan graphs

Eigenvalues, random walks and Ramanujan graphs Eigenvalues, random walks and Ramanujan graphs David Ellis 1 The Expander Mixing lemma We have seen that a bounded-degree graph is a good edge-expander if and only if if has large spectral gap If G = (V,

More information

Geometric Mapping Properties of Semipositive Matrices

Geometric Mapping Properties of Semipositive Matrices Geometric Mapping Properties of Semipositive Matrices M. J. Tsatsomeros Mathematics Department Washington State University Pullman, WA 99164 (tsat@wsu.edu) July 14, 2015 Abstract Semipositive matrices

More information

Stochastic modelling of epidemic spread

Stochastic modelling of epidemic spread Stochastic modelling of epidemic spread Julien Arino Department of Mathematics University of Manitoba Winnipeg Julien Arino@umanitoba.ca 19 May 2012 1 Introduction 2 Stochastic processes 3 The SIS model

More information

Lecture 1: Brief Review on Stochastic Processes

Lecture 1: Brief Review on Stochastic Processes Lecture 1: Brief Review on Stochastic Processes A stochastic process is a collection of random variables {X t (s) : t T, s S}, where T is some index set and S is the common sample space of the random variables.

More information

Lecture 11: Introduction to Markov Chains. Copyright G. Caire (Sample Lectures) 321

Lecture 11: Introduction to Markov Chains. Copyright G. Caire (Sample Lectures) 321 Lecture 11: Introduction to Markov Chains Copyright G. Caire (Sample Lectures) 321 Discrete-time random processes A sequence of RVs indexed by a variable n 2 {0, 1, 2,...} forms a discretetime random process

More information

STEINER 2-DESIGNS S(2, 4, 28) WITH NONTRIVIAL AUTOMORPHISMS. Vedran Krčadinac Department of Mathematics, University of Zagreb, Croatia

STEINER 2-DESIGNS S(2, 4, 28) WITH NONTRIVIAL AUTOMORPHISMS. Vedran Krčadinac Department of Mathematics, University of Zagreb, Croatia GLASNIK MATEMATIČKI Vol. 37(57)(2002), 259 268 STEINER 2-DESIGNS S(2, 4, 28) WITH NONTRIVIAL AUTOMORPHISMS Vedran Krčadinac Department of Mathematics, University of Zagreb, Croatia Abstract. In this article

More information

k-protected VERTICES IN BINARY SEARCH TREES

k-protected VERTICES IN BINARY SEARCH TREES k-protected VERTICES IN BINARY SEARCH TREES MIKLÓS BÓNA Abstract. We show that for every k, the probability that a randomly selected vertex of a random binary search tree on n nodes is at distance k from

More information

Chapter 2: Linear Independence and Bases

Chapter 2: Linear Independence and Bases MATH20300: Linear Algebra 2 (2016 Chapter 2: Linear Independence and Bases 1 Linear Combinations and Spans Example 11 Consider the vector v (1, 1 R 2 What is the smallest subspace of (the real vector space

More information

On a problem of Bermond and Bollobás

On a problem of Bermond and Bollobás On a problem of Bermond and Bollobás arxiv:1803.07501v1 [math.co] 20 Mar 2018 Slobodan Filipovski University of Primorska, Koper, Slovenia slobodan.filipovski@famnit.upr.si Robert Jajcay Comenius University,

More information

The Nearest Doubly Stochastic Matrix to a Real Matrix with the same First Moment

The Nearest Doubly Stochastic Matrix to a Real Matrix with the same First Moment he Nearest Doubly Stochastic Matrix to a Real Matrix with the same First Moment William Glunt 1, homas L. Hayden 2 and Robert Reams 2 1 Department of Mathematics and Computer Science, Austin Peay State

More information

Exercise Solutions to Functional Analysis

Exercise Solutions to Functional Analysis Exercise Solutions to Functional Analysis Note: References refer to M. Schechter, Principles of Functional Analysis Exersize that. Let φ,..., φ n be an orthonormal set in a Hilbert space H. Show n f n

More information

The Terwilliger Algebras of Group Association Schemes

The Terwilliger Algebras of Group Association Schemes The Terwilliger Algebras of Group Association Schemes Eiichi Bannai Akihiro Munemasa The Terwilliger algebra of an association scheme was introduced by Paul Terwilliger [7] in order to study P-and Q-polynomial

More information

Notes taken by Graham Taylor. January 22, 2005

Notes taken by Graham Taylor. January 22, 2005 CSC4 - Linear Programming and Combinatorial Optimization Lecture : Different forms of LP. The algebraic objects behind LP. Basic Feasible Solutions Notes taken by Graham Taylor January, 5 Summary: We first

More information

Peter J. Dukes. 22 August, 2012

Peter J. Dukes. 22 August, 2012 22 August, 22 Graph decomposition Let G and H be graphs on m n vertices. A decompostion of G into copies of H is a collection {H i } of subgraphs of G such that each H i = H, and every edge of G belongs

More information

Online solution of the average cost Kullback-Leibler optimization problem

Online solution of the average cost Kullback-Leibler optimization problem Online solution of the average cost Kullback-Leibler optimization problem Joris Bierkens Radboud University Nijmegen j.bierkens@science.ru.nl Bert Kappen Radboud University Nijmegen b.kappen@science.ru.nl

More information

Vector Spaces ปร ภ ม เวกเตอร

Vector Spaces ปร ภ ม เวกเตอร Vector Spaces ปร ภ ม เวกเตอร 5.1 Real Vector Spaces ปร ภ ม เวกเตอร ของจ านวนจร ง Vector Space Axioms (1/2) Let V be an arbitrary nonempty set of objects on which two operations are defined, addition and

More information

Solving the Hamiltonian Cycle problem using symbolic determinants

Solving the Hamiltonian Cycle problem using symbolic determinants Solving the Hamiltonian Cycle problem using symbolic determinants V. Ejov, J.A. Filar, S.K. Lucas & J.L. Nelson Abstract In this note we show how the Hamiltonian Cycle problem can be reduced to solving

More information

Classical Complexity and Fixed-Parameter Tractability of Simultaneous Consecutive Ones Submatrix & Editing Problems

Classical Complexity and Fixed-Parameter Tractability of Simultaneous Consecutive Ones Submatrix & Editing Problems Classical Complexity and Fixed-Parameter Tractability of Simultaneous Consecutive Ones Submatrix & Editing Problems Rani M. R, Mohith Jagalmohanan, R. Subashini Binary matrices having simultaneous consecutive

More information

An estimation of the spectral radius of a product of block matrices

An estimation of the spectral radius of a product of block matrices Linear Algebra and its Applications 379 (2004) 267 275 wwwelseviercom/locate/laa An estimation of the spectral radius of a product of block matrices Mei-Qin Chen a Xiezhang Li b a Department of Mathematics

More information

On Backward Product of Stochastic Matrices

On Backward Product of Stochastic Matrices On Backward Product of Stochastic Matrices Behrouz Touri and Angelia Nedić 1 Abstract We study the ergodicity of backward product of stochastic and doubly stochastic matrices by introducing the concept

More information

SCHUR IDEALS AND HOMOMORPHISMS OF THE SEMIDEFINITE CONE

SCHUR IDEALS AND HOMOMORPHISMS OF THE SEMIDEFINITE CONE SCHUR IDEALS AND HOMOMORPHISMS OF THE SEMIDEFINITE CONE BABHRU JOSHI AND M. SEETHARAMA GOWDA Abstract. We consider the semidefinite cone K n consisting of all n n real symmetric positive semidefinite matrices.

More information