CSC 576: Linear System
|
|
- Brittany Anthony
- 5 years ago
- Views:
Transcription
1 CSC 576: Linear System Ji Liu Department of Computer Science, University of Rochester September 3, 206 Linear Equations Consider solving linear equations where A R m n and b R n m and n could be extremely large 2 Preliminary Ax = b () σ i (A) denotes the ith largest singular value of A, σ min (A) denotes the minimal nonzero singular value of A and σ max (A) denotes the maximal singular value of A; If U has orthogonal columns, that is, U T U = I, then Ux = x and U y y ; Ax σ max (A) x, where σ max (A) denotes the largest singular value of A; Note that we do NOT have in general Ax σ min (A) x, where σ min (A)(> 0) denotes the minimal nonzero singular value of A However, we have AA T x σ min (A) A T x (due to AA T x = UΣ 2 U T x = Σ 2 U T x σ min (Σ) ΣU T x = σ min (Σ) V ΣU T x = σ min (A) A T x ); The compact SVD of A is A = UΣV T We have span(a) = span(u) and span(a T ) = span(v ); Let a and b two random variables we have E a,b [f(a, b)] = E b [E a b [f(a, b)]] = E a [E b a [f(a, b)]] The derivative (gradient, differential) of a function f(x) in terms of a matrix X R m n is defined below f(x) f(x) f(x) X X j X n f(x) X = f(x) f(x) f(x) X i X ij X in f(x) f(x) X m X mj f(x) X mn
2 The commonly used derivatives include f(x) = trace(a T f(x) X) = A, X X = A f(x) = trace(x T f(x) AX) X = A + AT f(x) = 2 AX f(x) 2 X = AT (AX ) f(x) = 2 trace(t X T f(x) X) X = XT f(x) = 2 trace(t X T f(x) AX) X = 2 (A + AT )X T A general rule to compute the derivative of f(x, X) is f(x, X) = f(x, X 2 ) + f(x, X 2 ) X X X 2 X=X2=X 3 Closed orm of Eq () 3 Case : A is invertible irst we consider that A is invertible (which implies that A a square matrix and has a full rank) The closed form solution is x = A b (2) Computational complexity: The complexity to compute the inverse of a matrix is O(n 3 ) The Gaussian elimination algorithm can achieve this complexity The memory requirement is O(n 2 ) Although Eq (2) gives the exact solution to problem (), it would be a disaster when n is huge This method is only used in solving small scale problem 32 Case 2: A may be not invertible If A might be not invertible, then there might not exist a solution or exist more than one solution to Eq () or example, A = [ 0; 0 ; 0 0], b = [; ; ], and A = [ 0 0; 0 0], b = [; ] In the more general situation, people use pseudo inverse to compute x: x = A + b where A + is defined in the following Denote the compact SVD of A as A = UΣV T The pseudo inverse of A is defined as A + = (UΣ V T ) T = V Σ U T Note that if A is invertible, then A = A + Now you may ask why is x = A + b? Actually, A + b solves the following objective: min x 2 Ax b 2 2
3 To verify this, we only need to check the optimality condition Substituting x = A + b into both sides, we have A T Ax = A T b A T A(A + b) = (UΣV T ) T (UΣV T )(V Σ U T )b = V ΣU T b = A T b It is worth to note that A(A + b) is the projection of b onto the range space of A Question: If there are multiple solutions to (), then A + b is just one of solutions So why this solution is so speical? Since A is just a special case of A +, the complexity of computing A + is still in the order of O(n 3 ) in general The next question is how to solve () when n is extremely large? Read next section! 4 (Randomized) Kaczmarz Algorithm irst we assume that Eq () has at least one solution (Even if this condition fails, we can reformulate it into a problem satisfying this condition, which will be clear soon) Note that we do not need assume that A is a square matrix The randomized Kaczmarz (RK) algorithm is a storage efficient algorithm Algorithm Randomized Kaczmarz Algorithm : Given A R m n and b R m ; 2: Initialize k 0 and x 0 = 0; 3: while k K do 4: Choose i from {, 2,, m} with probability { A i 2 / A 2 }; 5: Update 6: k k + ; 7: end while x k+ x k A i x k b i A i 2 A T i ; Intuition: The RK algorithm iteratively randomly selects a hyperplane A i x = b i and projects the current x onto this selected hyperplane Given a hyperplane c x = d, the projection of a given point z onto this hyperplane is proj {x c x=d}(z) = z c x d c 2 c which is obtained by considering two conditions: ) the difference z proj {x c x=d}(z) must be along the normal direction of the hyperplane, so it should be proportional to c; 2) the factor in front of c can be calculated from the observation that proj {x c x=d}(z) must be on the hyperplane Computational complexity: The RK algorithm only needs a row of the data matrix A for computation oth the memory cost and computation complexity are just O(n) Convergence rate: The RK algorithm is nothing but the stochastic gradient algorithm, which will be clear in our later class However, RK guarantees a much faster convergence rate then the general convergence rate of the stochastic gradient algorithm 3
4 Theorem Assume that Eq () has at least a solution Denote the minimal nonzero singular value of A as σ min (A) ( E( x k x 2 ) σ2 min (A) ) k A 2 x 0 x 0 2 where x = A + b and E means taking the expectation in terms of all random variables Proof irst we verify that x = A + b is a solution to Ax = b As we showed before, x essentially minimizes Ax b 2 Since there exists at least one solution to Ax = b, that is, min x Ax b 2 = 0, which implies that Ax = b We notice that x k (for any k) is in the span of all rows in A, that is, x k span{a T, AT 2,, AT m } or x k can be written as Ay k for some vector y k We also notice that x = A + b span{a T, AT 2,, AT m } Therefore x k x span{a T, AT 2,, AT m } or x k x = A T z k for some z k Using the result in our preliminary section, we have A(x k x ) = AA T z k σ min (A) Az k = σ min (A) (x k x ) Next we have x k+ x 2 = x k A i 2 AT i (A i x k b i ) x = x k x 2 + A i 4 AT i (A i x k b i ) 2 2 x k x, A T i (A i x k b i ) = x k x 2 + A i 2 A i x k b i 2 2 A i (x k x ), A i x k b i = x k x 2 A i 2 [A i (x k x )] 2 (from Ax = b) Taking expectation on both sides in term of i(k) given i(0), i(),, i(k ), we have E i(k) i(0), i(k ) ( x k+ x 2 ) = x k x k 2 E i(k) {i(0), i(k )} A i 2 [A i (x k x )] 2 which implies that = x k x k 2 i 2 A i 2 A 2 = x k x k 2 A 2 A(x k x ) 2 x k x k 2 σ2 min (A) x k x 2 A 2 ( = σ2 min (A) ) A 2 x k x 2, A i 2 [A i (x k x )] 2 E i(k),i(k ),,i(0) ( x k+ x 2 ( ) =E i(0), i(k ) Ei(k) i(0), i(k ) ( x k+ x 2 ) ) (( E i(0), i(k ) σ2 min (A) ) ) A 2 x k x 2 ( σ2 min (A) ) k+ A 2 x 0 x 2 4
5 It completes the proof If problem () does not have a solution, one usually solves or equivalently min x Ax b 2 2 A T Ax = A T b (3) One can directly apply RK algorithm to solve Eq (3) However, in many cases, it is hard to compute A T A in advance due to memory issue One way to deal with it is to solve any equivalent problem by introducing a dual variable y R m : A T y = A T b Ax = y or [ 0 A T A I ] [ ] x = y [ A T ] b 0 Question: How to solve the following linear system using the RK algorithm where A R m n, R k l, and C R n k AX = C 5 Conjugate Gradient Algorithm The conjugate gradient (CG) algorithm is one of the most important algorithms for solving linear systems However, the original paper which proposes this algorithm was rejected for many times, until people realized its merits CG aims at solving the following problem x = c where is a PSD matrix To solve a least squares problem 2 Ax b 2, one can apply CG to solve A T Ax = A T b (4) The CG algorithm is provided in Algorithm 2 To see the motivation of CG, let us consider the projection in the inner product space with inner product definition x, y = x y where is a positive definite matrix Assume that we have a group of orthogonal basis {p, p 2,, p n } which satisfy p i, p j = 0 for any i j Let x be the solution to x = c Apparently, we can find a unique decompose for x by x = n α i p i i= It implies that b = n α i p i i= 5
6 To decide the value of α i, we can use α i = b, p i p i 2 where we use the orthogonality among all basis, and p i 2 is defined as p i, p i Therefore, when we have the basis, the solution can be easily obtained So the key question is how to obtain the group of basis Next we introduce a general approach to construct a group of basis Given a group of independent vectors {r, r 2,, r n }, we can construct a group of orthogonal basis following the procedure below: p =r p 2 =r 2 r 2, p p 2 p p n =r n k<n r n, p k p k 2 p k The key motivation above is to remove the components p, p 2,, p k from r k to obtain p k One can verify p i, p j = 0 for any i j Now we know how to construct orthogonal basis in the space associating with However, the computational complexity is too high Can we do better by reducing the computational complexity? If we can find a group linear independent vectors {r k } n k= satisfy the following property then we can construct {p k } n k= by r k, p j = 0 j =, 2,, k 2, (5) p =r p 2 =r 2 r 2, p p 2 p p n =r n r n, p n p n 2 p n CG basically designs a smart way to construct {r k } n k= r =b, p = r r 2 =r r, p p 2 p = r b, p p 2 p p 2 = r 2 r 2, p p 2 p r n =r n r n, p n p n 2 p n = r n b, p n p n 2 p n, p n = r n r n, p n p n 2 p n 6
7 To see why (5) holds for the construction above, defining K j = span(b, b,, j b) we have r K p K r 2 K 2 r 2 K p 2 K 2 r 3 K 3 r 3 K 2 p 3 K 3 r n K n r n K n p n K n rom p j K j+, we obtain that r k, p j = 0 for all j k 2 Algorithm 2 Conjugate Gradient : r 0 = A T b A T Ax 0 ; 2: p 0 = r 0 ; 3: k = 0; 4: while k K do 5: Compute the steplength α k = r k 2 Ap k 2 6: Update x k+ = x k + α k p k 7: r k+ = r k α k A T Ap k 8: if r k+ is sufficiently small then exit loop 9: β k = r k+ 2 r k 2 0: Compute the conjugate gradient p k+ = r k+ + β k p k : k k + 2: end while Theorem 2 Apply CG to (4) The kth iterate of CG algorithm satisfies ( ( x k x 2 O σ ) ) k min(a) x 0 x 2 σ max (A) The computational complexity of CG per iteration is O(mn) To compare RK and CG, we assume that rank(r) = m Then we only need to compare ( ) m σ2 m(a) A 2 σ2 m(a) m A 2 which suggests that CG outperforms RK if = m σm(a) 2 r i= σ2 i (A) and σ m(a) σ (A), σ m (A)σ (A) m m σi 2 (A) i= 6 Successive over-relaxation Consider the linear system () with a full rank A 7
8 Ax = b where a a 2 a n x b a 2 a 22 a 2n A =, x = x 2, b = b 2 a n a n2 a nn x n b n Then A can be decomposed into a diagonal matrix diagonal component D, and triangular matrix Strictly triangular matrix strictly lower and upper triangular components L and U : A = D + L + U, where a a 2 a n 0 a 22 0 D =, L = a 2 0 0, U = 0 0 a 2n 0 0 a nn a n a n The system of linear equations may be rewritten as: (D + ωl)x = ωb [ωu + (ω )D]x for a constant ω >, called the relaxation factor The method of successive over-relaxation is an Iterative method iterative technique that solves the left hand side of this expression for x, using previous value for x on the right hand side Analytically, this may be written as: x (k+) = (D + ωl) ( ωb [ωu + (ω )D]x (k)) = L w x (k) + c However, by taking advantage of the triangular form of (D + ωl), the elements of x k+) can be computed sequentially using [[forward substitution]]: x (i) k+ = ( ω)x(i) k+ + ω a ii b i j<i a ij x (j) k+ j>i a ij x (j) k, i =, 2,, n The choice of relaxation factor ω is not necessarily easy, and depends upon the properties of the coefficient matrix In 947, Ostrowski proved that if A is Symmetric matrix symmetric and Positive-definite matrix positive-definite then ρ(l ω ) < for 0 < ω < 2 Thus convergence of the iteration process follows, but we are generally interested in faster convergence rather than just convergence 8
Chapter 7 Iterative Techniques in Matrix Algebra
Chapter 7 Iterative Techniques in Matrix Algebra Per-Olof Persson persson@berkeley.edu Department of Mathematics University of California, Berkeley Math 128B Numerical Analysis Vector Norms Definition
More informationComputational Methods. Systems of Linear Equations
Computational Methods Systems of Linear Equations Manfred Huber 2010 1 Systems of Equations Often a system model contains multiple variables (parameters) and contains multiple equations Multiple equations
More informationBasic Elements of Linear Algebra
A Basic Review of Linear Algebra Nick West nickwest@stanfordedu September 16, 2010 Part I Basic Elements of Linear Algebra Although the subject of linear algebra is much broader than just vectors and matrices,
More informationAM 205: lecture 8. Last time: Cholesky factorization, QR factorization Today: how to compute the QR factorization, the Singular Value Decomposition
AM 205: lecture 8 Last time: Cholesky factorization, QR factorization Today: how to compute the QR factorization, the Singular Value Decomposition QR Factorization A matrix A R m n, m n, can be factorized
More informationLecture Note 7: Iterative methods for solving linear systems. Xiaoqun Zhang Shanghai Jiao Tong University
Lecture Note 7: Iterative methods for solving linear systems Xiaoqun Zhang Shanghai Jiao Tong University Last updated: December 24, 2014 1.1 Review on linear algebra Norms of vectors and matrices vector
More informationApplied Mathematics 205. Unit II: Numerical Linear Algebra. Lecturer: Dr. David Knezevic
Applied Mathematics 205 Unit II: Numerical Linear Algebra Lecturer: Dr. David Knezevic Unit II: Numerical Linear Algebra Chapter II.3: QR Factorization, SVD 2 / 66 QR Factorization 3 / 66 QR Factorization
More informationThe QR Factorization
The QR Factorization How to Make Matrices Nicer Radu Trîmbiţaş Babeş-Bolyai University March 11, 2009 Radu Trîmbiţaş ( Babeş-Bolyai University) The QR Factorization March 11, 2009 1 / 25 Projectors A projector
More informationA Brief Outline of Math 355
A Brief Outline of Math 355 Lecture 1 The geometry of linear equations; elimination with matrices A system of m linear equations with n unknowns can be thought of geometrically as m hyperplanes intersecting
More informationLinear Algebra and Matrix Inversion
Jim Lambers MAT 46/56 Spring Semester 29- Lecture 2 Notes These notes correspond to Section 63 in the text Linear Algebra and Matrix Inversion Vector Spaces and Linear Transformations Matrices are much
More informationLinear Algebra Primer
Linear Algebra Primer David Doria daviddoria@gmail.com Wednesday 3 rd December, 2008 Contents Why is it called Linear Algebra? 4 2 What is a Matrix? 4 2. Input and Output.....................................
More informationConjugate Gradient algorithm. Storage: fixed, independent of number of steps.
Conjugate Gradient algorithm Need: A symmetric positive definite; Cost: 1 matrix-vector product per step; Storage: fixed, independent of number of steps. The CG method minimizes the A norm of the error,
More informationThe Singular Value Decomposition
The Singular Value Decomposition Philippe B. Laval KSU Fall 2015 Philippe B. Laval (KSU) SVD Fall 2015 1 / 13 Review of Key Concepts We review some key definitions and results about matrices that will
More informationLeast Squares. Tom Lyche. October 26, Centre of Mathematics for Applications, Department of Informatics, University of Oslo
Least Squares Tom Lyche Centre of Mathematics for Applications, Department of Informatics, University of Oslo October 26, 2010 Linear system Linear system Ax = b, A C m,n, b C m, x C n. under-determined
More informationMatrix decompositions
Matrix decompositions Zdeněk Dvořák May 19, 2015 Lemma 1 (Schur decomposition). If A is a symmetric real matrix, then there exists an orthogonal matrix Q and a diagonal matrix D such that A = QDQ T. The
More informationEE263: Introduction to Linear Dynamical Systems Review Session 9
EE63: Introduction to Linear Dynamical Systems Review Session 9 SVD continued EE63 RS9 1 Singular Value Decomposition recall any nonzero matrix A R m n, with Rank(A) = r, has an SVD given by A = UΣV T,
More informationCS412: Lecture #17. Mridul Aanjaneya. March 19, 2015
CS: Lecture #7 Mridul Aanjaneya March 9, 5 Solving linear systems of equations Consider a lower triangular matrix L: l l l L = l 3 l 3 l 33 l n l nn A procedure similar to that for upper triangular systems
More informationIterative techniques in matrix algebra
Iterative techniques in matrix algebra Tsung-Ming Huang Department of Mathematics National Taiwan Normal University, Taiwan September 12, 2015 Outline 1 Norms of vectors and matrices 2 Eigenvalues and
More informationOrthogonalization and least squares methods
Chapter 3 Orthogonalization and least squares methods 31 QR-factorization (QR-decomposition) 311 Householder transformation Definition 311 A complex m n-matrix R = [r ij is called an upper (lower) triangular
More informationIterative Methods for Solving A x = b
Iterative Methods for Solving A x = b A good (free) online source for iterative methods for solving A x = b is given in the description of a set of iterative solvers called templates found at netlib: http
More informationMAT 2037 LINEAR ALGEBRA I web:
MAT 237 LINEAR ALGEBRA I 2625 Dokuz Eylül University, Faculty of Science, Department of Mathematics web: Instructor: Engin Mermut http://kisideuedutr/enginmermut/ HOMEWORK 2 MATRIX ALGEBRA Textbook: Linear
More informationB553 Lecture 5: Matrix Algebra Review
B553 Lecture 5: Matrix Algebra Review Kris Hauser January 19, 2012 We have seen in prior lectures how vectors represent points in R n and gradients of functions. Matrices represent linear transformations
More informationApplications of Randomized Methods for Decomposing and Simulating from Large Covariance Matrices
Applications of Randomized Methods for Decomposing and Simulating from Large Covariance Matrices Vahid Dehdari and Clayton V. Deutsch Geostatistical modeling involves many variables and many locations.
More informationCOURSE Iterative methods for solving linear systems
COURSE 0 4.3. Iterative methods for solving linear systems Because of round-off errors, direct methods become less efficient than iterative methods for large systems (>00 000 variables). An iterative scheme
More informationLINEAR ALGEBRA SUMMARY SHEET.
LINEAR ALGEBRA SUMMARY SHEET RADON ROSBOROUGH https://intuitiveexplanationscom/linear-algebra-summary-sheet/ This document is a concise collection of many of the important theorems of linear algebra, organized
More informationThe 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 informationPreliminary/Qualifying Exam in Numerical Analysis (Math 502a) Spring 2012
Instructions Preliminary/Qualifying Exam in Numerical Analysis (Math 502a) Spring 2012 The exam consists of four problems, each having multiple parts. You should attempt to solve all four problems. 1.
More informationThe Gram Schmidt Process
u 2 u The Gram Schmidt Process Now we will present a procedure, based on orthogonal projection, that converts any linearly independent set of vectors into an orthogonal set. Let us begin with the simple
More informationThe Gram Schmidt Process
The Gram Schmidt Process Now we will present a procedure, based on orthogonal projection, that converts any linearly independent set of vectors into an orthogonal set. Let us begin with the simple case
More informationNumerical Linear Algebra Primer. Ryan Tibshirani Convex Optimization /36-725
Numerical Linear Algebra Primer Ryan Tibshirani Convex Optimization 10-725/36-725 Last time: proximal gradient descent Consider the problem min g(x) + h(x) with g, h convex, g differentiable, and h simple
More informationSOLVING SPARSE LINEAR SYSTEMS OF EQUATIONS. Chao Yang Computational Research Division Lawrence Berkeley National Laboratory Berkeley, CA, USA
1 SOLVING SPARSE LINEAR SYSTEMS OF EQUATIONS Chao Yang Computational Research Division Lawrence Berkeley National Laboratory Berkeley, CA, USA 2 OUTLINE Sparse matrix storage format Basic factorization
More information. =. a i1 x 1 + a i2 x 2 + a in x n = b i. a 11 a 12 a 1n a 21 a 22 a 1n. i1 a i2 a in
Vectors and Matrices Continued Remember that our goal is to write a system of algebraic equations as a matrix equation. Suppose we have the n linear algebraic equations a x + a 2 x 2 + a n x n = b a 2
More informationProperties 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 information1 Multiply Eq. E i by λ 0: (λe i ) (E i ) 2 Multiply Eq. E j by λ and add to Eq. E i : (E i + λe j ) (E i )
Direct Methods for Linear Systems Chapter Direct Methods for Solving Linear Systems Per-Olof Persson persson@berkeleyedu Department of Mathematics University of California, Berkeley Math 18A Numerical
More informationLecture 18 Classical Iterative Methods
Lecture 18 Classical Iterative Methods MIT 18.335J / 6.337J Introduction to Numerical Methods Per-Olof Persson November 14, 2006 1 Iterative Methods for Linear Systems Direct methods for solving Ax = b,
More informationSTA141C: Big Data & High Performance Statistical Computing
STA141C: Big Data & High Performance Statistical Computing Numerical Linear Algebra Background Cho-Jui Hsieh UC Davis May 15, 2018 Linear Algebra Background Vectors A vector has a direction and a magnitude
More informationThis can be accomplished by left matrix multiplication as follows: I
1 Numerical Linear Algebra 11 The LU Factorization Recall from linear algebra that Gaussian elimination is a method for solving linear systems of the form Ax = b, where A R m n and bran(a) In this method
More informationSTA141C: Big Data & High Performance Statistical Computing
STA141C: Big Data & High Performance Statistical Computing Lecture 5: Numerical Linear Algebra Cho-Jui Hsieh UC Davis April 20, 2017 Linear Algebra Background Vectors A vector has a direction and a magnitude
More informationLinear Algebra. Session 12
Linear Algebra. Session 12 Dr. Marco A Roque Sol 08/01/2017 Example 12.1 Find the constant function that is the least squares fit to the following data x 0 1 2 3 f(x) 1 0 1 2 Solution c = 1 c = 0 f (x)
More informationLINEAR SYSTEMS (11) Intensive Computation
LINEAR SYSTEMS () Intensive Computation 27-8 prof. Annalisa Massini Viviana Arrigoni EXACT METHODS:. GAUSSIAN ELIMINATION. 2. CHOLESKY DECOMPOSITION. ITERATIVE METHODS:. JACOBI. 2. GAUSS-SEIDEL 2 CHOLESKY
More informationLinear Analysis Lecture 16
Linear Analysis Lecture 16 The QR Factorization Recall the Gram-Schmidt orthogonalization process. Let V be an inner product space, and suppose a 1,..., a n V are linearly independent. Define q 1,...,
More informationLecture 9: Numerical Linear Algebra Primer (February 11st)
10-725/36-725: Convex Optimization Spring 2015 Lecture 9: Numerical Linear Algebra Primer (February 11st) Lecturer: Ryan Tibshirani Scribes: Avinash Siravuru, Guofan Wu, Maosheng Liu Note: LaTeX template
More informationJim Lambers MAT 610 Summer Session Lecture 1 Notes
Jim Lambers MAT 60 Summer Session 2009-0 Lecture Notes Introduction This course is about numerical linear algebra, which is the study of the approximate solution of fundamental problems from linear algebra
More informationThe University of Texas at Austin Department of Electrical and Computer Engineering. EE381V: Large Scale Learning Spring 2013.
The University of Texas at Austin Department of Electrical and Computer Engineering EE381V: Large Scale Learning Spring 2013 Assignment Two Caramanis/Sanghavi Due: Tuesday, Feb. 19, 2013. Computational
More informationMATH 3511 Lecture 1. Solving Linear Systems 1
MATH 3511 Lecture 1 Solving Linear Systems 1 Dmitriy Leykekhman Spring 2012 Goals Review of basic linear algebra Solution of simple linear systems Gaussian elimination D Leykekhman - MATH 3511 Introduction
More informationLinear Algebra Review. Vectors
Linear Algebra Review 9/4/7 Linear Algebra Review By Tim K. Marks UCSD Borrows heavily from: Jana Kosecka http://cs.gmu.edu/~kosecka/cs682.html Virginia de Sa (UCSD) Cogsci 8F Linear Algebra review Vectors
More informationProcess Model Formulation and Solution, 3E4
Process Model Formulation and Solution, 3E4 Section B: Linear Algebraic Equations Instructor: Kevin Dunn dunnkg@mcmasterca Department of Chemical Engineering Course notes: Dr Benoît Chachuat 06 October
More informationCOMP 558 lecture 18 Nov. 15, 2010
Least squares We have seen several least squares problems thus far, and we will see more in the upcoming lectures. For this reason it is good to have a more general picture of these problems and how to
More informationMaths for Signals and Systems Linear Algebra in Engineering
Maths for Signals and Systems Linear Algebra in Engineering Lectures 13 15, Tuesday 8 th and Friday 11 th November 016 DR TANIA STATHAKI READER (ASSOCIATE PROFFESOR) IN SIGNAL PROCESSING IMPERIAL COLLEGE
More information5. Orthogonal matrices
L Vandenberghe EE133A (Spring 2017) 5 Orthogonal matrices matrices with orthonormal columns orthogonal matrices tall matrices with orthonormal columns complex matrices with orthonormal columns 5-1 Orthonormal
More informationLinear Algebra, Summer 2011, pt. 2
Linear Algebra, Summer 2, pt. 2 June 8, 2 Contents Inverses. 2 Vector Spaces. 3 2. Examples of vector spaces..................... 3 2.2 The column space......................... 6 2.3 The null space...........................
More informationG1110 & 852G1 Numerical Linear Algebra
The University of Sussex Department of Mathematics G & 85G Numerical Linear Algebra Lecture Notes Autumn Term Kerstin Hesse (w aw S w a w w (w aw H(wa = (w aw + w Figure : Geometric explanation of the
More informationAlgorithms to Compute Bases and the Rank of a Matrix
Algorithms to Compute Bases and the Rank of a Matrix Subspaces associated to a matrix Suppose that A is an m n matrix The row space of A is the subspace of R n spanned by the rows of A The column space
More informationYORK UNIVERSITY. Faculty of Science Department of Mathematics and Statistics MATH M Test #2 Solutions
YORK UNIVERSITY Faculty of Science Department of Mathematics and Statistics MATH 3. M Test # Solutions. (8 pts) For each statement indicate whether it is always TRUE or sometimes FALSE. Note: For this
More informationChapter 4 No. 4.0 Answer True or False to the following. Give reasons for your answers.
MATH 434/534 Theoretical Assignment 3 Solution Chapter 4 No 40 Answer True or False to the following Give reasons for your answers If a backward stable algorithm is applied to a computational problem,
More informationChapter 7: Symmetric Matrices and Quadratic Forms
Chapter 7: Symmetric Matrices and Quadratic Forms (Last Updated: December, 06) These notes are derived primarily from Linear Algebra and its applications by David Lay (4ed). A few theorems have been moved
More informationAMS526: Numerical Analysis I (Numerical Linear Algebra)
AMS526: Numerical Analysis I (Numerical Linear Algebra) Lecture 5: Projectors and QR Factorization Xiangmin Jiao SUNY Stony Brook Xiangmin Jiao Numerical Analysis I 1 / 14 Outline 1 Projectors 2 QR Factorization
More informationPrincipal Component Analysis
Machine Learning Michaelmas 2017 James Worrell Principal Component Analysis 1 Introduction 1.1 Goals of PCA Principal components analysis (PCA) is a dimensionality reduction technique that can be used
More informationNORMS ON SPACE OF MATRICES
NORMS ON SPACE OF MATRICES. Operator Norms on Space of linear maps Let A be an n n real matrix and x 0 be a vector in R n. We would like to use the Picard iteration method to solve for the following system
More informationSection 6.4. The Gram Schmidt Process
Section 6.4 The Gram Schmidt Process Motivation The procedures in 6 start with an orthogonal basis {u, u,..., u m}. Find the B-coordinates of a vector x using dot products: x = m i= x u i u i u i u i Find
More informationMATH 240 Spring, Chapter 1: Linear Equations and Matrices
MATH 240 Spring, 2006 Chapter Summaries for Kolman / Hill, Elementary Linear Algebra, 8th Ed. Sections 1.1 1.6, 2.1 2.2, 3.2 3.8, 4.3 4.5, 5.1 5.3, 5.5, 6.1 6.5, 7.1 7.2, 7.4 DEFINITIONS Chapter 1: Linear
More informationSingular Value Decomposition
Singular Value Decomposition Motivatation The diagonalization theorem play a part in many interesting applications. Unfortunately not all matrices can be factored as A = PDP However a factorization A =
More informationThe Singular Value Decomposition
The Singular Value Decomposition We are interested in more than just sym+def matrices. But the eigenvalue decompositions discussed in the last section of notes will play a major role in solving general
More informationAMS526: Numerical Analysis I (Numerical Linear Algebra for Computational and Data Sciences)
AMS526: Numerical Analysis I (Numerical Linear Algebra for Computational and Data Sciences) Lecture 19: Computing the SVD; Sparse Linear Systems Xiangmin Jiao Stony Brook University Xiangmin Jiao Numerical
More informationVector Spaces, Orthogonality, and Linear Least Squares
Week Vector Spaces, Orthogonality, and Linear Least Squares. Opening Remarks.. Visualizing Planes, Lines, and Solutions Consider the following system of linear equations from the opener for Week 9: χ χ
More informationOrthonormal Transformations and Least Squares
Orthonormal Transformations and Least Squares Tom Lyche Centre of Mathematics for Applications, Department of Informatics, University of Oslo October 30, 2009 Applications of Qx with Q T Q = I 1. solving
More informationReview of Basic Concepts in Linear Algebra
Review of Basic Concepts in Linear Algebra Grady B Wright Department of Mathematics Boise State University September 7, 2017 Math 565 Linear Algebra Review September 7, 2017 1 / 40 Numerical Linear Algebra
More informationx 3y 2z = 6 1.2) 2x 4y 3z = 8 3x + 6y + 8z = 5 x + 3y 2z + 5t = 4 1.5) 2x + 8y z + 9t = 9 3x + 5y 12z + 17t = 7
Linear Algebra and its Applications-Lab 1 1) Use Gaussian elimination to solve the following systems x 1 + x 2 2x 3 + 4x 4 = 5 1.1) 2x 1 + 2x 2 3x 3 + x 4 = 3 3x 1 + 3x 2 4x 3 2x 4 = 1 x + y + 2z = 4 1.4)
More information2. Linear algebra. matrices and vectors. linear equations. range and nullspace of matrices. function of vectors, gradient and Hessian
FE661 - Statistical Methods for Financial Engineering 2. Linear algebra Jitkomut Songsiri matrices and vectors linear equations range and nullspace of matrices function of vectors, gradient and Hessian
More informationAlgebra C Numerical Linear Algebra Sample Exam Problems
Algebra C Numerical Linear Algebra Sample Exam Problems Notation. Denote by V a finite-dimensional Hilbert space with inner product (, ) and corresponding norm. The abbreviation SPD is used for symmetric
More informationBlock Bidiagonal Decomposition and Least Squares Problems
Block Bidiagonal Decomposition and Least Squares Problems Åke Björck Department of Mathematics Linköping University Perspectives in Numerical Analysis, Helsinki, May 27 29, 2008 Outline Bidiagonal Decomposition
More informationInverse Singular Value Problems
Chapter 8 Inverse Singular Value Problems IEP versus ISVP Existence question A continuous approach An iterative method for the IEP An iterative method for the ISVP 139 140 Lecture 8 IEP versus ISVP Inverse
More informationMATH 315 Linear Algebra Homework #1 Assigned: August 20, 2018
Homework #1 Assigned: August 20, 2018 Review the following subjects involving systems of equations and matrices from Calculus II. Linear systems of equations Converting systems to matrix form Pivot entry
More informationNotes on Linear Algebra
1 Notes on Linear Algebra Jean Walrand August 2005 I INTRODUCTION Linear Algebra is the theory of linear transformations Applications abound in estimation control and Markov chains You should be familiar
More informationCS137 Introduction to Scientific Computing Winter Quarter 2004 Solutions to Homework #3
CS137 Introduction to Scientific Computing Winter Quarter 2004 Solutions to Homework #3 Felix Kwok February 27, 2004 Written Problems 1. (Heath E3.10) Let B be an n n matrix, and assume that B is both
More informationMIT Final Exam Solutions, Spring 2017
MIT 8.6 Final Exam Solutions, Spring 7 Problem : For some real matrix A, the following vectors form a basis for its column space and null space: C(A) = span,, N(A) = span,,. (a) What is the size m n of
More informationIr O D = D = ( ) Section 2.6 Example 1. (Bottom of page 119) dim(v ) = dim(l(v, W )) = dim(v ) dim(f ) = dim(v )
Section 3.2 Theorem 3.6. Let A be an m n matrix of rank r. Then r m, r n, and, by means of a finite number of elementary row and column operations, A can be transformed into the matrix ( ) Ir O D = 1 O
More informationAMS526: Numerical Analysis I (Numerical Linear Algebra for Computational and Data Sciences)
AMS526: Numerical Analysis I (Numerical Linear Algebra for Computational and Data Sciences) Lecture 1: Course Overview; Matrix Multiplication Xiangmin Jiao Stony Brook University Xiangmin Jiao Numerical
More informationLinear Algebra, part 3. Going back to least squares. Mathematical Models, Analysis and Simulation = 0. a T 1 e. a T n e. Anna-Karin Tornberg
Linear Algebra, part 3 Anna-Karin Tornberg Mathematical Models, Analysis and Simulation Fall semester, 2010 Going back to least squares (Sections 1.7 and 2.3 from Strang). We know from before: The vector
More informationLinear Equations and Matrix
1/60 Chia-Ping Chen Professor Department of Computer Science and Engineering National Sun Yat-sen University Linear Algebra Gaussian Elimination 2/60 Alpha Go Linear algebra begins with a system of linear
More informationChapter 4 - MATRIX ALGEBRA. ... a 2j... a 2n. a i1 a i2... a ij... a in
Chapter 4 - MATRIX ALGEBRA 4.1. Matrix Operations A a 11 a 12... a 1j... a 1n a 21. a 22.... a 2j... a 2n. a i1 a i2... a ij... a in... a m1 a m2... a mj... a mn The entry in the ith row and the jth column
More informationTHE SINGULAR VALUE DECOMPOSITION MARKUS GRASMAIR
THE SINGULAR VALUE DECOMPOSITION MARKUS GRASMAIR 1. Definition Existence Theorem 1. Assume that A R m n. Then there exist orthogonal matrices U R m m V R n n, values σ 1 σ 2... σ p 0 with p = min{m, n},
More informationMODULE 8 Topics: Null space, range, column space, row space and rank of a matrix
MODULE 8 Topics: Null space, range, column space, row space and rank of a matrix Definition: Let L : V 1 V 2 be a linear operator. The null space N (L) of L is the subspace of V 1 defined by N (L) = {x
More informationLinear Systems of n equations for n unknowns
Linear Systems of n equations for n unknowns In many application problems we want to find n unknowns, and we have n linear equations Example: Find x,x,x such that the following three equations hold: x
More informationMATH 583A REVIEW SESSION #1
MATH 583A REVIEW SESSION #1 BOJAN DURICKOVIC 1. Vector Spaces Very quick review of the basic linear algebra concepts (see any linear algebra textbook): (finite dimensional) vector space (or linear space),
More informationNumerical Linear Algebra Primer. Ryan Tibshirani Convex Optimization
Numerical Linear Algebra Primer Ryan Tibshirani Convex Optimization 10-725 Consider Last time: proximal Newton method min x g(x) + h(x) where g, h convex, g twice differentiable, and h simple. Proximal
More informationThe Singular Value Decomposition and Least Squares Problems
The Singular Value Decomposition and Least Squares Problems Tom Lyche Centre of Mathematics for Applications, Department of Informatics, University of Oslo September 27, 2009 Applications of SVD solving
More informationComputational Methods CMSC/AMSC/MAPL 460. EigenValue decomposition Singular Value Decomposition. Ramani Duraiswami, Dept. of Computer Science
Computational Methods CMSC/AMSC/MAPL 460 EigenValue decomposition Singular Value Decomposition Ramani Duraiswami, Dept. of Computer Science Hermitian Matrices A square matrix for which A = A H is said
More informationThroughout these notes we assume V, W are finite dimensional inner product spaces over C.
Math 342 - Linear Algebra II Notes Throughout these notes we assume V, W are finite dimensional inner product spaces over C 1 Upper Triangular Representation Proposition: Let T L(V ) There exists an orthonormal
More informationDS-GA 1002 Lecture notes 0 Fall Linear Algebra. These notes provide a review of basic concepts in linear algebra.
DS-GA 1002 Lecture notes 0 Fall 2016 Linear Algebra These notes provide a review of basic concepts in linear algebra. 1 Vector spaces You are no doubt familiar with vectors in R 2 or R 3, i.e. [ ] 1.1
More informationThe conjugate gradient method
The conjugate gradient method Michael S. Floater November 1, 2011 These notes try to provide motivation and an explanation of the CG method. 1 The method of conjugate directions We want to solve the linear
More informationECE133A Applied Numerical Computing Additional Lecture Notes
Winter Quarter 2018 ECE133A Applied Numerical Computing Additional Lecture Notes L. Vandenberghe ii Contents 1 LU factorization 1 1.1 Definition................................. 1 1.2 Nonsingular sets
More informationBindel, Fall 2016 Matrix Computations (CS 6210) Notes for
1 Iteration basics Notes for 2016-11-07 An iterative solver for Ax = b is produces a sequence of approximations x (k) x. We always stop after finitely many steps, based on some convergence criterion, e.g.
More informationEE731 Lecture Notes: Matrix Computations for Signal Processing
EE731 Lecture Notes: Matrix Computations for Signal Processing James P. Reilly c Department of Electrical and Computer Engineering McMaster University September 22, 2005 0 Preface This collection of ten
More informationChapter 3. Linear and Nonlinear Systems
59 An expert is someone who knows some of the worst mistakes that can be made in his subject, and how to avoid them Werner Heisenberg (1901-1976) Chapter 3 Linear and Nonlinear Systems In this chapter
More informationFundamentals of Engineering Analysis (650163)
Philadelphia University Faculty of Engineering Communications and Electronics Engineering Fundamentals of Engineering Analysis (6563) Part Dr. Omar R Daoud Matrices: Introduction DEFINITION A matrix is
More informationIterative Solution methods
p. 1/28 TDB NLA Parallel Algorithms for Scientific Computing Iterative Solution methods p. 2/28 TDB NLA Parallel Algorithms for Scientific Computing Basic Iterative Solution methods The ideas to use iterative
More informationbe a Householder matrix. Then prove the followings H = I 2 uut Hu = (I 2 uu u T u )u = u 2 uut u
MATH 434/534 Theoretical Assignment 7 Solution Chapter 7 (71) Let H = I 2uuT Hu = u (ii) Hv = v if = 0 be a Householder matrix Then prove the followings H = I 2 uut Hu = (I 2 uu )u = u 2 uut u = u 2u =
More informationFall TMA4145 Linear Methods. Exercise set Given the matrix 1 2
Norwegian University of Science and Technology Department of Mathematical Sciences TMA445 Linear Methods Fall 07 Exercise set Please justify your answers! The most important part is how you arrive at an
More informationComputational Economics and Finance
Computational Economics and Finance Part II: Linear Equations Spring 2016 Outline Back Substitution, LU and other decomposi- Direct methods: tions Error analysis and condition numbers Iterative methods:
More informationCHAPTER 6. Direct Methods for Solving Linear Systems
CHAPTER 6 Direct Methods for Solving Linear Systems. Introduction A direct method for approximating the solution of a system of n linear equations in n unknowns is one that gives the exact solution to
More information