Nonlinear Overlapping Domain Decomposition Methods

Size: px
Start display at page:

Download "Nonlinear Overlapping Domain Decomposition Methods"

Transcription

1 Nonlnear Overlappng Doman Decomposton Methods Xao-Chuan Ca 1 Department of Computer Scence, Unversty of Colorado at Boulder, Boulder, CO 80309, ca@cs.colorado.edu Summary. We dscuss some overlappng doman decomposton algorthms for solvng sparse nonlnear system of equatons arsng from the dscretzaton of partal dfferental equatons. All algorthms are derved usng the three basc algorthms: Newton for local or global nonlnear systems, Krylov for the lnear Jacoban system nsde Newton, and Schwarz for lnear and/or nonlnear precondtonng. The two key ssues wth nonlnear solvers are robustness and parallel scalablty. Both ssues can be addressed f a good combnaton of Newton, Krylov and Schwarz s selected, and the rght selecton s often dependent on the partcular type of nonlnearty and the computng platform. 1 Introducton For solvng partal dfferental equatons on large scale parallel computers, doman decomposton s a natural choce. Overlappng Schwarz methods and non-overlappng teratve substructurng methods are the two major classes of doman decomposton methods [13, 14, 15]. In ths paper we only consder overlappng methods for solvng large sparse nonlnear system of equatons arsng from the dscretzaton of nonlnear partal dfferental equatons,.e., for a gven nonlnear functon F : R n R n, we compute a vector u R n, such that F (u) = 0, (1) startng from an ntal guess u (0) R n. Here F = (F 1,..., F n ) T, F = F (u 1,..., u n ), and u = (u 1,..., u n ) T. One of the popularly used technques for solvng (1) s the so-called nexact Newton algorthms (IN) whch are descrbed brefly here. Suppose u (k) s the current approxmate soluton and J = F (u (k) ), a new approxmate soluton u (k+1) can be computed through The research was supported n part by DOE under DE-FC02-01ER25479 and DE- FC02-04ER25595, and n part by NSF under grants ACI , CNS , CCF , and CNS

2 2 Xao-Chuan Ca the followng steps: frst fnd an nexact Newton drecton p (k) by solvng the Jacoban system Jp (k) = F (u (k) ) (2) such that F (u (k) ) Jp (k) η k F (u (k) ), then compute the new approxmate soluton u (k+1) = u (k) λ (k) p (k). (3) Here η k [0, 1) s a scalar that determnes how accurately the Jacoban system needs to be solved usng, for example, Krylov subspace methods. λ (k) s another scalar that determnes the step length n the selected nexact Newton drecton. Sometmes when J s not explctly avalable, one can use the matrxfree verson [11]. IN has several well-known features. (a) Fast convergence. If the ntal guess s close enough to the desred soluton then the convergence s very fast (quadratc) provded that the η k s are suffcently small. (b) Non-robustness. The convergence, or fast convergence, happens only f a good ntal guess s avalable. Generally t s dffcult to obtan such an ntal guess especally for nonlnear equatons that have unbalanced nonlneartes [12]. The step length λ (k) s often determned by the components wth the strongest nonlneartes, and ths may lead to an extended perod of stagnaton n the nonlnear resdual curve. We say that the nonlneartes are unbalanced when λ (k), n effect, s determned by a subset of the overall degrees of freedom. (c) Scalablty. The parallel scalablty of the method s mostly determned by how the Jacoban system (2) s solved. There are a number of strateges [7, 8, 10], such as lnesearch, trust regon, contnuaton or better ways to choose the forcng term, to make the algorthm more robust or converge faster, however, these strateges are all based on certan global knowledge of F or J. In other words, all equatons n the system are treated equally as f they were some of the worst equatons n the system. Other ways to look at the global nature of IN are (d) To advance from u (k) to u (k+1), all n varables and equatons need to be updated even though n many stuatons n can be very large, but only a small number of components of u (k) receve sgnfcant updates. (e) If a small number of components of the ntal guess u (0) are not acceptable, the entre u (0) s declared bad. (f) There are two global control varables η k and λ (k). Any slght change of F may result n the change of η k or λ (k), and any slght change of η k or λ (k) may result n some global functon evaluatons and/or the solvng of global Jacoban systems. For example, f the search drecton p (k) has one unacceptable component, then the entre steplength s reduced. Note that these global operatons can be expensve when n s large and when the number of processors s large. Usng doman decomposton methods, more

3 Nonlnear Doman Decomposton Methods 3 localzed treatments can be appled based on the locaton or the physcal nature of the nonlneartes, and the number of global operatons can be made small n some stuatons. We should pont out that the words local and global have dfferent meanngs n the context of doman decomposton methods [15] than n the context of nonlnear equaton solvers [7], among others. In nonlnear solvers, local means a small neghborhood of the exact soluton of the nonlnear system, and global means a relatvely large neghborhood of the exact soluton of the nonlnear system. In doman decomposton, local means some subregons n the computatonal doman and global means the whole computatonal doman. All the algorthms to be dscussed n the paper are constructed wth a combnaton of the three basc technques: Newton, Krylov and Schwarz. Newton s the basc nonlnear solver that s used for ether the system defned on the whole space or some subspaces (subdoman subspace or coarse subspace). Krylov s the basc lnear solver that s used nsde a Newton solver. Schwarz s a precondtoner for ether the lnear or the nonlnear solver. Many algorthms can be derved wth dfferent combnatons of the three basc algorthms. For a gven class of problems and computng platform, a specal combnaton mght be necessary n order to obtan the best performance. The three basc algorthms are all well understood ndvdually, however, the constructon of the best combnaton remans a challenge. The same can be sad for the software. All software components are readly avalable n PETSc [1], but some of the advanced combnatons have to be programmed by the user. We next defne (nformally) some notatons for descrbng doman decomposton methods. u s understood as a dscrete (or coeffcents of a fnte element) functon defned on the computatonal doman Ω whch s already parttoned nto a set of subdomans {Ω1, δ, ΩN δ }. Here Ωδ s an δ-extenson of Ω, and the collecton of {Ω } s a non-overlappng partton of Ω. We defne R δ as a restrcton operator assocated wth Ωδ and R0 as the restrcton operator assocated wth Ω. We denote u Ω δ as the restrcton of u on Ω δ, and u Ω δ as the restrcton of u on the boundary of Ω δ. Here we use the word doman to denote the mesh ponts n the nteror of the doman and boundary to denote the mesh ponts on the boundary of the doman. Smlarly, we may restrct the nonlnear functon to a subdoman, such as. For boundary value problems consdered n ths paper, we assume (u) = (u Ω δ, u Ω δ ). That s to say that there are no global equatons n the system that may couple the equatons defned at a mesh pont to equatons defned outsde a small neghborhood. The rest of the paper s organzed as follows. In Secton 2, we dscuss the most popular overlappng nonlnear doman decomposton method, Newton- Krylov-Schwarz algorthm, and n Sectons 3-6, we dscuss some more advanced nonlnear methods. Some fnal remarks are gven n Secton 7.

4 4 Xao-Chuan Ca 2 Newton-Krylov-Schwarz algorthms Newton-Krylov-Schwarz (NKS) s smply the applcaton of a lnear Schwarz precondtoner for solvng the Jacoban equaton (2) n the nexact Newton algorthm [2, 3]. Dependng on what type of Schwarz precondtoner s used (addtve, multplcatve, restrcted, one-level, two-level, etc), there are several NKS algorthms. Let us defne the subdoman precondtoners as J = R δ J(R δ ) T, = 1,..., N, then the addtve Schwarz precondtoner can be wrtten as M 1 N AS = (R δ ) T J 1 R δ. =1 Because of ts smplcty, NKS has become one of the most popular doman decomposton methods for solvng nonlnear PDEs and s the default nonlnear solver n PETSc [1]. The nonlnear propertes of NKS are exactly the same as that of nexact Newton. For example, the ntal guess has to be suffcently close to the soluton n order to obtan convergence, and fast convergence can be acheved when the nonlnearty s well balanced. NKS addresses the scalablty ssue (c) of IN well, but not the other ssues (a, b, d-f). 3 The classcal Schwarz alternatng algorthms Let (u (0),..., u (0) ) be the ntal guess for all subdomans. The classcal Ω1 δ ΩN δ Schwarz alternatng algorthm (SA) can be descrbed as follows: k = 1,..., tll convergence condton s satsfed = 1,..., N defne u (k) usng {u (k 1), 1 j N} or {u (k) Ω δ Ωj δ Ωj δ compute u (k) Ω δ by solvng (u (k), u (k) ) = 0. Ω δ Ω δ, 1 j < } The algorthm doesn t belong to the class of IN algorthms and, n general, not share propertes (a-f). The method s usually not used by tself as a nonlnear solver because of ts slow convergence, but n some cases when the nonlneartes are solated wthn some of the subdomans, the method can be a good alternatve to IN. Note that SA doesn t nvolve any global operatons. 4 Nonlnear addtve Schwarz precondtoned nexact Newton algorthms The basc dea of nonlnearly precondtoned nexact Newton algorthms ([4, 9]) s to fnd the soluton u R n of (1) by solvng an equvalent system

5 Nonlnear Doman Decomposton Methods 5 F(u) = 0 (4) usng IN. Systems (1) and (4) are sad to be equvalent f they have the same soluton. For any gven v R n, we defne a subdoman projecton T (v), whch s a functon wth support n Ω δ, as the soluton of the followng subspace nonlnear system (v T (v)) = 0, for = 1,..., N. Then a nonlnearly precondtoned functon s defned as F(u) = N T (u). =1 It can be shown that, under certan condtons, for ths partcular F, (1) and (4) offer the same soluton subject to the error due to dfferent stoppng condtons and precondtoners. Ths algorthm s often referred to as the addtve Schwarz precondtoned nexact Newton algorthm (ASPIN). Sometmes we call t a left precondtoned IN because n the lnear case (.e., F (u) = Ju b) F(u) = ( N =1 (Rδ )T J 1 R δ )(Ju b). When usng IN to solve (4), the Jacoban of F, or ts approxmaton, s needed. Because of the specal defnton of the functon F, ts Jacoban can only be gven as the sum of matrx-vector products and the explct elements of F are not avalable. It s known that for left precondtoned lnear teratve methods, the stoppng condton s often nfluenced by the precondtoner. The mpact of the precondtoner on the stoppng condton can be removed f the precondtoner s appled to the rght. Unlke lnear precondtonng, the swtch from left to rght s not trval n the nonlnear case. A rght nonlnear precondtoner wll be dscuss n a later secton of the paper. 5 Nonlnear elmnaton algorthms The nonlnear elmnaton algorthm (NE) was ntroduced n [12] for nonlnear algebrac systems wth local hgh nonlneartes. It was not ntroduced as a doman decomposton method, but we nclude t n the paper because t s the man motvaton for the algorthm to be dscussed n the next secton. Suppose that the functon F s more nonlnear n the subdoman Ω δ, then we can elmnate all unknowns n ths partcular subdoman and let Newton work on the rest of the varables and equatons. Let y = u Ω δ and x = u Ω\Ω δ, then usng the mplct functon theorem, under some assumptons, we can solve for y n terms of x;.e., solve (x, y) = 0

6 6 Xao-Chuan Ca for y, whch symbolcally equals to y = F 1 (x). After the elmnaton, we Ω δ can use the regular Newton method for the rest of the system whch s more balanced, at least n theory, ( ) F Ω\Ω δ x, F 1 (x) = 0. Ω δ The algorthm has some obvous advantages. We menton some of ts dsadvantages as a motvaton for the algorthm to be dscussed n the next secton. In practce, t s often dffcult to tell whch components are more nonlnear than the others, and the stuaton may change from teraton to teraton. The algorthm may ntroduce sharp jumps n the resdual functon near the nterface of x and y. Such jumps may lead to slow convergence or dvergence. Some mproved versons are gven n [6]. In the next secton, we combne the deas of ASPIN and NE nto a rght precondtoned Newton method. 6 Nonlnear restrcted addtve Schwarz algorthms In [5], a rght precondtoned nexact Newton algorthm was ntroduced as follows: Fnd the soluton u R n of (1) by frst solvng a precondtoned nonlnear system F (G(v)) = 0 for v, and then obtan u = G(v). For any gven v R n, we defne a subdoman projecton T (v), whch s a functon wth support n Ω δ, as the soluton of the followng subspace nonlnear system (v + T (v)) = 0, for = 1,..., N. Then the nonlnear precondtonng functon s defned as G(v) = v + N R 0 T (v). =1 Here the non-overlappng restrcton operator R 0 effectvely removes the sharp jumps on the nterfaces of the overlappng subdomans. In the lnear case ( N ) G(v) = v (R 0 ) T J 1 R δ (Jv b), =1 whch can be regarded as a restrcted addtve Schwarz precondtoned Rchardson method. Ths precondtoner doesn t have to be appled at every outer Newton teraton. It s used only when some local hgh nonlneartes are sensed, somehow. Below we descrbe the overall algorthm (NKS-RAS). The goal s to solve equaton (1) wth a gven ntal guess u (0). Suppose u (k) s the current soluton.

7 Nonlnear Doman Decomposton Methods 7 Step 1 (The Nonlnearty Checkng Step): Check local and global stoppng condtons. If the global condton s satsfed, stop. If local condtons ndcate that nonlneartes are not balanced, go to Step 2. If local condtons ndcate that nonlneartes are balanced, set ũ (k) = u (k), go to Step 3. Step 2 (The RAS Step): Solve local nonlnear problems on the overlappng subdomans to obtan the subdoman correctons T (u (k) ) ) (u (k) + T (u (k) ) = 0, for = 1,, N. Drop the soluton n the overlappng part of the subdoman and compute the global functon G(u (k) ) and set ( ũ (k) = G u (k)). Go to Step 3. Step 3 (The NKS Step): Compute the next approxmate soluton u (k+1) by solvng the followng system F (u) = 0 wth one step of NKS usng ũ (k) as the ntal guess. Go to Step 1. The nonlnearty checkng step s mportant. However, we only have a few ad hoc technques such as computng the resdual norm subdoman by subdoman (or feld by feld n the case of mult-physcs applcatons). If some of the subdoman (or sub-feld) norms are much larger than for other subdomans, we label these subdomans as hghly nonlnear subdomans and proceed wth the RAS elmnaton step. Otherwse, when the nonlnearty s more or less balanced we bypass the RAS step and go drectly to the global NKS step. The subdoman nonlnear systems n Step 2 do not need to be solved very accurately snce the solutons are used only to construct an ntal guess for Step 3. In NKS-RAS, a nonlnear system s set up on each subdoman, but n practce, not all subdoman nonlnear problem needs to be solved. In the nottoo-nonlnear regons, the solver may declare to have converged n 0 teraton. 7 Concludng remarks In ths paper, we have gven a quck overvew of overlappng doman decomposton methods for solvng nonlnear partal dfferental equatons. The two key ssues of nonlnear methods are robustness and scalablty. Both ssues

8 8 Xao-Chuan Ca can be addressed by usng some combnatons of the three basc algorthms: Newton, Krylov and Schwarz. Several algorthms are presented n the paper together wth some of ther advantages and dsadvantages. Dependng on the partcular types of nonlneartes and the computng platform, dfferent combnatons of the three basc algorthms may be needed n order to obtan the best performance and robustness. Due to page lmt, applcatons have not been dscussed n the paper. Some of them can be found n the references. References 1. S. Balay, K. Buschelman, W. D. Gropp, D. Kaushk, M. Knepley, L. C. McInnes, B. F. Smth, and H. Zhang, PETSc Users Manual, ANL, X.-C. Ca, W. Gropp, D. Keyes, R. Melvn, and D. P. Young, Parallel Newton- Krylov-Schwarz algorthms for the transonc full potental equaton, SIAM J. Sc. Comput., 19 (1998), pp X.-C. Ca, W. D. Gropp, D. E. Keyes, and M. D. Tdrr, Newton-Krylov- Schwarz methods n CFD, Proceedngs of the Internatonal Workshop on the Naver-Stokes Equatons, Notes n Numercal Flud Mechancs, R. Rannacher, eds. Veweg Verlag, Braunschweg, X.-C. Ca and D. E. Keyes, Nonlnearly precondtoned nexact Newton algorthms, SIAM J. Sc. Comput., 24 (2002), pp X.-C. Ca and X. L, Inexact Newton methods wth nonlnear restrcted addtve Schwarz precondtonng for problems wth hgh local nonlneartes, n preparaton 6. X.-C. Ca and X. L, A doman decomposton based parallel nexact Newtons method wth subspace correcton for ncompressble Naver-Stokes equatons, Lecture Notes n Computer Scence, Sprnger, J. E. Denns and R. B. Schnabel, Numercal Methods for Unconstraned Optmzaton and Nonlnear Equatons, SIAM, S. C. Esenstat and H. F. Walker, Choosng the forcng terms n an nexact Newton method, SIAM J. Sc. Comput., 17 (1996), pp F.-N. Hwang and X.-C. Ca, A parallel nonlnear addtve Schwarz precondtoned nexact Newton algorthm for ncompressble Naver-Stokes equatons, J. Comput. Phys., 204 (2005), pp C. T. Kelley and D. E. Keyes, Convergence analyss of pseudo-transent contnuaton, SIAM J. Numer. Anal., 35 (1998), pp D. Knoll and D. E. Keyes, Jacoban-free Newton-Krylov methods: a survey of approaches and applcatons, J. Comput. Phys., 193 (2004), pp P. J. Lanzkron, D. J. Rose, and J. T. Wlkes, An analyss of approxmate nonlnear elmnaton, SIAM J. Sc. Comput., 17 (1996), pp B. Smth, P. Bjørstad, and W. Gropp, Doman Decomposton. Parallel Multlevel Methods for Elptc Partal Dfferental Equatons, Cambrdge Unversty Press, New York, A. Quarteron and A. Vall, Doman Decomposton Methods for Partal Dfferental Equatons, Oxford Unversty Press, Oxford, A. Tosell and O. Wdlund, Doman Decomposton Methods Algorthms and Theory, Sprnger, Berln, 2005

NONLINEARLY PRECONDITIONED INEXACT NEWTON ALGORITHMS

NONLINEARLY PRECONDITIONED INEXACT NEWTON ALGORITHMS NONLINEARLY PRECONDITIONED INEXACT NEWTON ALGORITHMS XIAO-CHUAN CAI AND DAVID E. KEYES Abstract. Inexact Newton algorthms are commonly used for solvng large sparse nonlnear system of equatons F (u ) =

More information

Inexact Newton Methods for Inverse Eigenvalue Problems

Inexact Newton Methods for Inverse Eigenvalue Problems Inexact Newton Methods for Inverse Egenvalue Problems Zheng-jan Ba Abstract In ths paper, we survey some of the latest development n usng nexact Newton-lke methods for solvng nverse egenvalue problems.

More information

Numerical Heat and Mass Transfer

Numerical Heat and Mass Transfer Master degree n Mechancal Engneerng Numercal Heat and Mass Transfer 06-Fnte-Dfference Method (One-dmensonal, steady state heat conducton) Fausto Arpno f.arpno@uncas.t Introducton Why we use models and

More information

CHAPTER 5 NUMERICAL EVALUATION OF DYNAMIC RESPONSE

CHAPTER 5 NUMERICAL EVALUATION OF DYNAMIC RESPONSE CHAPTER 5 NUMERICAL EVALUATION OF DYNAMIC RESPONSE Analytcal soluton s usually not possble when exctaton vares arbtrarly wth tme or f the system s nonlnear. Such problems can be solved by numercal tmesteppng

More information

Deriving the X-Z Identity from Auxiliary Space Method

Deriving the X-Z Identity from Auxiliary Space Method Dervng the X-Z Identty from Auxlary Space Method Long Chen Department of Mathematcs, Unversty of Calforna at Irvne, Irvne, CA 92697 chenlong@math.uc.edu 1 Iteratve Methods In ths paper we dscuss teratve

More information

Hongyi Miao, College of Science, Nanjing Forestry University, Nanjing ,China. (Received 20 June 2013, accepted 11 March 2014) I)ϕ (k)

Hongyi Miao, College of Science, Nanjing Forestry University, Nanjing ,China. (Received 20 June 2013, accepted 11 March 2014) I)ϕ (k) ISSN 1749-3889 (prnt), 1749-3897 (onlne) Internatonal Journal of Nonlnear Scence Vol.17(2014) No.2,pp.188-192 Modfed Block Jacob-Davdson Method for Solvng Large Sparse Egenproblems Hongy Mao, College of

More information

Lecture 12: Discrete Laplacian

Lecture 12: Discrete Laplacian Lecture 12: Dscrete Laplacan Scrbe: Tanye Lu Our goal s to come up wth a dscrete verson of Laplacan operator for trangulated surfaces, so that we can use t n practce to solve related problems We are mostly

More information

Chapter 5. Solution of System of Linear Equations. Module No. 6. Solution of Inconsistent and Ill Conditioned Systems

Chapter 5. Solution of System of Linear Equations. Module No. 6. Solution of Inconsistent and Ill Conditioned Systems Numercal Analyss by Dr. Anta Pal Assstant Professor Department of Mathematcs Natonal Insttute of Technology Durgapur Durgapur-713209 emal: anta.bue@gmal.com 1 . Chapter 5 Soluton of System of Lnear Equatons

More information

A Hybrid Variational Iteration Method for Blasius Equation

A Hybrid Variational Iteration Method for Blasius Equation Avalable at http://pvamu.edu/aam Appl. Appl. Math. ISSN: 1932-9466 Vol. 10, Issue 1 (June 2015), pp. 223-229 Applcatons and Appled Mathematcs: An Internatonal Journal (AAM) A Hybrd Varatonal Iteraton Method

More information

Difference Equations

Difference Equations Dfference Equatons c Jan Vrbk 1 Bascs Suppose a sequence of numbers, say a 0,a 1,a,a 3,... s defned by a certan general relatonshp between, say, three consecutve values of the sequence, e.g. a + +3a +1

More information

Appendix B. The Finite Difference Scheme

Appendix B. The Finite Difference Scheme 140 APPENDIXES Appendx B. The Fnte Dfference Scheme In ths appendx we present numercal technques whch are used to approxmate solutons of system 3.1 3.3. A comprehensve treatment of theoretcal and mplementaton

More information

On a direct solver for linear least squares problems

On a direct solver for linear least squares problems ISSN 2066-6594 Ann. Acad. Rom. Sc. Ser. Math. Appl. Vol. 8, No. 2/2016 On a drect solver for lnear least squares problems Constantn Popa Abstract The Null Space (NS) algorthm s a drect solver for lnear

More information

MMA and GCMMA two methods for nonlinear optimization

MMA and GCMMA two methods for nonlinear optimization MMA and GCMMA two methods for nonlnear optmzaton Krster Svanberg Optmzaton and Systems Theory, KTH, Stockholm, Sweden. krlle@math.kth.se Ths note descrbes the algorthms used n the author s 2007 mplementatons

More information

Lectures - Week 4 Matrix norms, Conditioning, Vector Spaces, Linear Independence, Spanning sets and Basis, Null space and Range of a Matrix

Lectures - Week 4 Matrix norms, Conditioning, Vector Spaces, Linear Independence, Spanning sets and Basis, Null space and Range of a Matrix Lectures - Week 4 Matrx norms, Condtonng, Vector Spaces, Lnear Independence, Spannng sets and Bass, Null space and Range of a Matrx Matrx Norms Now we turn to assocatng a number to each matrx. We could

More information

A New Refinement of Jacobi Method for Solution of Linear System Equations AX=b

A New Refinement of Jacobi Method for Solution of Linear System Equations AX=b Int J Contemp Math Scences, Vol 3, 28, no 17, 819-827 A New Refnement of Jacob Method for Soluton of Lnear System Equatons AX=b F Naem Dafchah Department of Mathematcs, Faculty of Scences Unversty of Gulan,

More information

Non shape regular domain decompositions: an analysis using a stable decomposition in H 1 0

Non shape regular domain decompositions: an analysis using a stable decomposition in H 1 0 Non shape regular doman decompostons: an analyss usng a stable decomposton n H 1 0 Martn Gander 1, Laurence Halpern, and Kévn Santugn Repquet 3 Abstract In ths paper, we establsh the exstence of a stable

More information

Global Sensitivity. Tuesday 20 th February, 2018

Global Sensitivity. Tuesday 20 th February, 2018 Global Senstvty Tuesday 2 th February, 28 ) Local Senstvty Most senstvty analyses [] are based on local estmates of senstvty, typcally by expandng the response n a Taylor seres about some specfc values

More information

APPENDIX A Some Linear Algebra

APPENDIX A Some Linear Algebra APPENDIX A Some Lnear Algebra The collecton of m, n matrces A.1 Matrces a 1,1,..., a 1,n A = a m,1,..., a m,n wth real elements a,j s denoted by R m,n. If n = 1 then A s called a column vector. Smlarly,

More information

Generalized Linear Methods

Generalized Linear Methods Generalzed Lnear Methods 1 Introducton In the Ensemble Methods the general dea s that usng a combnaton of several weak learner one could make a better learner. More formally, assume that we have a set

More information

Transfer Functions. Convenient representation of a linear, dynamic model. A transfer function (TF) relates one input and one output: ( ) system

Transfer Functions. Convenient representation of a linear, dynamic model. A transfer function (TF) relates one input and one output: ( ) system Transfer Functons Convenent representaton of a lnear, dynamc model. A transfer functon (TF) relates one nput and one output: x t X s y t system Y s The followng termnology s used: x y nput output forcng

More information

STAT 309: MATHEMATICAL COMPUTATIONS I FALL 2018 LECTURE 16

STAT 309: MATHEMATICAL COMPUTATIONS I FALL 2018 LECTURE 16 STAT 39: MATHEMATICAL COMPUTATIONS I FALL 218 LECTURE 16 1 why teratve methods f we have a lnear system Ax = b where A s very, very large but s ether sparse or structured (eg, banded, Toepltz, banded plus

More information

Errors for Linear Systems

Errors for Linear Systems Errors for Lnear Systems When we solve a lnear system Ax b we often do not know A and b exactly, but have only approxmatons  and ˆb avalable. Then the best thng we can do s to solve ˆx ˆb exactly whch

More information

Vector Norms. Chapter 7 Iterative Techniques in Matrix Algebra. Cauchy-Bunyakovsky-Schwarz Inequality for Sums. Distances. Convergence.

Vector Norms. Chapter 7 Iterative Techniques in Matrix Algebra. Cauchy-Bunyakovsky-Schwarz Inequality for Sums. Distances. Convergence. Vector Norms Chapter 7 Iteratve Technques n Matrx Algebra Per-Olof Persson persson@berkeley.edu Department of Mathematcs Unversty of Calforna, Berkeley Math 128B Numercal Analyss Defnton A vector norm

More information

n α j x j = 0 j=1 has a nontrivial solution. Here A is the n k matrix whose jth column is the vector for all t j=0

n α j x j = 0 j=1 has a nontrivial solution. Here A is the n k matrix whose jth column is the vector for all t j=0 MODULE 2 Topcs: Lnear ndependence, bass and dmenson We have seen that f n a set of vectors one vector s a lnear combnaton of the remanng vectors n the set then the span of the set s unchanged f that vector

More information

Report on Image warping

Report on Image warping Report on Image warpng Xuan Ne, Dec. 20, 2004 Ths document summarzed the algorthms of our mage warpng soluton for further study, and there s a detaled descrpton about the mplementaton of these algorthms.

More information

Lecture 21: Numerical methods for pricing American type derivatives

Lecture 21: Numerical methods for pricing American type derivatives Lecture 21: Numercal methods for prcng Amercan type dervatves Xaoguang Wang STAT 598W Aprl 10th, 2014 (STAT 598W) Lecture 21 1 / 26 Outlne 1 Fnte Dfference Method Explct Method Penalty Method (STAT 598W)

More information

Inner Product. Euclidean Space. Orthonormal Basis. Orthogonal

Inner Product. Euclidean Space. Orthonormal Basis. Orthogonal Inner Product Defnton 1 () A Eucldean space s a fnte-dmensonal vector space over the reals R, wth an nner product,. Defnton 2 (Inner Product) An nner product, on a real vector space X s a symmetrc, blnear,

More information

Overlapping additive and multiplicative Schwarz iterations for H -matrices

Overlapping additive and multiplicative Schwarz iterations for H -matrices Lnear Algebra and ts Applcatons 393 (2004) 91 105 www.elsever.com/locate/laa Overlappng addtve and multplcatve Schwarz teratons for H -matrces Rafael Bru a,1, Francsco Pedroche a, Danel B. Szyld b,,2 a

More information

Some Comments on Accelerating Convergence of Iterative Sequences Using Direct Inversion of the Iterative Subspace (DIIS)

Some Comments on Accelerating Convergence of Iterative Sequences Using Direct Inversion of the Iterative Subspace (DIIS) Some Comments on Acceleratng Convergence of Iteratve Sequences Usng Drect Inverson of the Iteratve Subspace (DIIS) C. Davd Sherrll School of Chemstry and Bochemstry Georga Insttute of Technology May 1998

More information

Introduction to Vapor/Liquid Equilibrium, part 2. Raoult s Law:

Introduction to Vapor/Liquid Equilibrium, part 2. Raoult s Law: CE304, Sprng 2004 Lecture 4 Introducton to Vapor/Lqud Equlbrum, part 2 Raoult s Law: The smplest model that allows us do VLE calculatons s obtaned when we assume that the vapor phase s an deal gas, and

More information

NUMERICAL DIFFERENTIATION

NUMERICAL DIFFERENTIATION NUMERICAL DIFFERENTIATION 1 Introducton Dfferentaton s a method to compute the rate at whch a dependent output y changes wth respect to the change n the ndependent nput x. Ths rate of change s called the

More information

Module 3 LOSSY IMAGE COMPRESSION SYSTEMS. Version 2 ECE IIT, Kharagpur

Module 3 LOSSY IMAGE COMPRESSION SYSTEMS. Version 2 ECE IIT, Kharagpur Module 3 LOSSY IMAGE COMPRESSION SYSTEMS Verson ECE IIT, Kharagpur Lesson 6 Theory of Quantzaton Verson ECE IIT, Kharagpur Instructonal Objectves At the end of ths lesson, the students should be able to:

More information

Econ107 Applied Econometrics Topic 3: Classical Model (Studenmund, Chapter 4)

Econ107 Applied Econometrics Topic 3: Classical Model (Studenmund, Chapter 4) I. Classcal Assumptons Econ7 Appled Econometrcs Topc 3: Classcal Model (Studenmund, Chapter 4) We have defned OLS and studed some algebrac propertes of OLS. In ths topc we wll study statstcal propertes

More information

NON-CENTRAL 7-POINT FORMULA IN THE METHOD OF LINES FOR PARABOLIC AND BURGERS' EQUATIONS

NON-CENTRAL 7-POINT FORMULA IN THE METHOD OF LINES FOR PARABOLIC AND BURGERS' EQUATIONS IJRRAS 8 (3 September 011 www.arpapress.com/volumes/vol8issue3/ijrras_8_3_08.pdf NON-CENTRAL 7-POINT FORMULA IN THE METHOD OF LINES FOR PARABOLIC AND BURGERS' EQUATIONS H.O. Bakodah Dept. of Mathematc

More information

RESTRICTED ADDITIVE SCHWARZ METHOD WITH HARMONIC OVERLAP

RESTRICTED ADDITIVE SCHWARZ METHOD WITH HARMONIC OVERLAP RESTRICTED ADDITIVE SCHWARZ METHOD WITH HARMONIC OVERLAP XIAO-CHUAN CAI, MAKSYMILIAN DRYJA, AND MARCUS SARKIS Abstract. In ths paper, we ntroduce a new Schwarz precondtoner and wth a new coarse space.

More information

For now, let us focus on a specific model of neurons. These are simplified from reality but can achieve remarkable results.

For now, let us focus on a specific model of neurons. These are simplified from reality but can achieve remarkable results. Neural Networks : Dervaton compled by Alvn Wan from Professor Jtendra Malk s lecture Ths type of computaton s called deep learnng and s the most popular method for many problems, such as computer vson

More information

A new Approach for Solving Linear Ordinary Differential Equations

A new Approach for Solving Linear Ordinary Differential Equations , ISSN 974-57X (Onlne), ISSN 974-5718 (Prnt), Vol. ; Issue No. 1; Year 14, Copyrght 13-14 by CESER PUBLICATIONS A new Approach for Solvng Lnear Ordnary Dfferental Equatons Fawz Abdelwahd Department of

More information

Feature Selection: Part 1

Feature Selection: Part 1 CSE 546: Machne Learnng Lecture 5 Feature Selecton: Part 1 Instructor: Sham Kakade 1 Regresson n the hgh dmensonal settng How do we learn when the number of features d s greater than the sample sze n?

More information

Robust Norm Equivalencies and Preconditioning

Robust Norm Equivalencies and Preconditioning Robust Norm Equvalences and Precondtonng Karl Scherer Insttut für Angewandte Mathematk, Unversty of Bonn, Wegelerstr. 6, 53115 Bonn, Germany Summary. In ths contrbuton we report on work done n contnuaton

More information

Relaxation Methods for Iterative Solution to Linear Systems of Equations

Relaxation Methods for Iterative Solution to Linear Systems of Equations Relaxaton Methods for Iteratve Soluton to Lnear Systems of Equatons Gerald Recktenwald Portland State Unversty Mechancal Engneerng Department gerry@pdx.edu Overvew Techncal topcs Basc Concepts Statonary

More information

A MODIFIED METHOD FOR SOLVING SYSTEM OF NONLINEAR EQUATIONS

A MODIFIED METHOD FOR SOLVING SYSTEM OF NONLINEAR EQUATIONS Journal of Mathematcs and Statstcs 9 (1): 4-8, 1 ISSN 1549-644 1 Scence Publcatons do:1.844/jmssp.1.4.8 Publshed Onlne 9 (1) 1 (http://www.thescpub.com/jmss.toc) A MODIFIED METHOD FOR SOLVING SYSTEM OF

More information

3.1 Expectation of Functions of Several Random Variables. )' be a k-dimensional discrete or continuous random vector, with joint PMF p (, E X E X1 E X

3.1 Expectation of Functions of Several Random Variables. )' be a k-dimensional discrete or continuous random vector, with joint PMF p (, E X E X1 E X Statstcs 1: Probablty Theory II 37 3 EPECTATION OF SEVERAL RANDOM VARIABLES As n Probablty Theory I, the nterest n most stuatons les not on the actual dstrbuton of a random vector, but rather on a number

More information

Yong Joon Ryang. 1. Introduction Consider the multicommodity transportation problem with convex quadratic cost function. 1 2 (x x0 ) T Q(x x 0 )

Yong Joon Ryang. 1. Introduction Consider the multicommodity transportation problem with convex quadratic cost function. 1 2 (x x0 ) T Q(x x 0 ) Kangweon-Kyungk Math. Jour. 4 1996), No. 1, pp. 7 16 AN ITERATIVE ROW-ACTION METHOD FOR MULTICOMMODITY TRANSPORTATION PROBLEMS Yong Joon Ryang Abstract. The optmzaton problems wth quadratc constrants often

More information

On an Extension of Stochastic Approximation EM Algorithm for Incomplete Data Problems. Vahid Tadayon 1

On an Extension of Stochastic Approximation EM Algorithm for Incomplete Data Problems. Vahid Tadayon 1 On an Extenson of Stochastc Approxmaton EM Algorthm for Incomplete Data Problems Vahd Tadayon Abstract: The Stochastc Approxmaton EM (SAEM algorthm, a varant stochastc approxmaton of EM, s a versatle tool

More information

Additional Codes using Finite Difference Method. 1 HJB Equation for Consumption-Saving Problem Without Uncertainty

Additional Codes using Finite Difference Method. 1 HJB Equation for Consumption-Saving Problem Without Uncertainty Addtonal Codes usng Fnte Dfference Method Benamn Moll 1 HJB Equaton for Consumpton-Savng Problem Wthout Uncertanty Before consderng the case wth stochastc ncome n http://www.prnceton.edu/~moll/ HACTproect/HACT_Numercal_Appendx.pdf,

More information

c 2006 Society for Industrial and Applied Mathematics

c 2006 Society for Industrial and Applied Mathematics SIAM J. MARIX ANAL. APPL. Vol. 29, No. 1, pp. 98 116 c 2006 Socety for Industral and Appled Mathematcs SCHWARZ IERAIONS FOR SYMMERIC POSIIVE SEMIDEFINIE PROBLEMS REINHARD NABBEN AND DANIEL B. SZYLD Abstract.

More information

EEE 241: Linear Systems

EEE 241: Linear Systems EEE : Lnear Systems Summary #: Backpropagaton BACKPROPAGATION The perceptron rule as well as the Wdrow Hoff learnng were desgned to tran sngle layer networks. They suffer from the same dsadvantage: they

More information

The Finite Element Method

The Finite Element Method The Fnte Element Method GENERAL INTRODUCTION Read: Chapters 1 and 2 CONTENTS Engneerng and analyss Smulaton of a physcal process Examples mathematcal model development Approxmate solutons and methods of

More information

CME 302: NUMERICAL LINEAR ALGEBRA FALL 2005/06 LECTURE 13

CME 302: NUMERICAL LINEAR ALGEBRA FALL 2005/06 LECTURE 13 CME 30: NUMERICAL LINEAR ALGEBRA FALL 005/06 LECTURE 13 GENE H GOLUB 1 Iteratve Methods Very large problems (naturally sparse, from applcatons): teratve methods Structured matrces (even sometmes dense,

More information

Handout # 6 (MEEN 617) Numerical Integration to Find Time Response of SDOF mechanical system. and write EOM (1) as two first-order Eqs.

Handout # 6 (MEEN 617) Numerical Integration to Find Time Response of SDOF mechanical system. and write EOM (1) as two first-order Eqs. Handout # 6 (MEEN 67) Numercal Integraton to Fnd Tme Response of SDOF mechancal system State Space Method The EOM for a lnear system s M X + DX + K X = F() t () t = X = X X = X = V wth ntal condtons, at

More information

Overlapping Domain Decomposition Methods for Linear Inverse Problems

Overlapping Domain Decomposition Methods for Linear Inverse Problems Overlappng Doman Decomposton Methods for Lnear Inverse Problems Daun Jang Hu Feng Jun Zou May, 4 Abstract We shall derve and propose several effcent overlappng doman decomposton methods for solvng some

More information

Some modelling aspects for the Matlab implementation of MMA

Some modelling aspects for the Matlab implementation of MMA Some modellng aspects for the Matlab mplementaton of MMA Krster Svanberg krlle@math.kth.se Optmzaton and Systems Theory Department of Mathematcs KTH, SE 10044 Stockholm September 2004 1. Consdered optmzaton

More information

Ballot Paths Avoiding Depth Zero Patterns

Ballot Paths Avoiding Depth Zero Patterns Ballot Paths Avodng Depth Zero Patterns Henrch Nederhausen and Shaun Sullvan Florda Atlantc Unversty, Boca Raton, Florda nederha@fauedu, ssull21@fauedu 1 Introducton In a paper by Sapounaks, Tasoulas,

More information

College of Computer & Information Science Fall 2009 Northeastern University 20 October 2009

College of Computer & Information Science Fall 2009 Northeastern University 20 October 2009 College of Computer & Informaton Scence Fall 2009 Northeastern Unversty 20 October 2009 CS7880: Algorthmc Power Tools Scrbe: Jan Wen and Laura Poplawsk Lecture Outlne: Prmal-dual schema Network Desgn:

More information

Linear Feature Engineering 11

Linear Feature Engineering 11 Lnear Feature Engneerng 11 2 Least-Squares 2.1 Smple least-squares Consder the followng dataset. We have a bunch of nputs x and correspondng outputs y. The partcular values n ths dataset are x y 0.23 0.19

More information

The Two-scale Finite Element Errors Analysis for One Class of Thermoelastic Problem in Periodic Composites

The Two-scale Finite Element Errors Analysis for One Class of Thermoelastic Problem in Periodic Composites 7 Asa-Pacfc Engneerng Technology Conference (APETC 7) ISBN: 978--6595-443- The Two-scale Fnte Element Errors Analyss for One Class of Thermoelastc Problem n Perodc Compostes Xaoun Deng Mngxang Deng ABSTRACT

More information

Lecture 13 APPROXIMATION OF SECOMD ORDER DERIVATIVES

Lecture 13 APPROXIMATION OF SECOMD ORDER DERIVATIVES COMPUTATIONAL FLUID DYNAMICS: FDM: Appromaton of Second Order Dervatves Lecture APPROXIMATION OF SECOMD ORDER DERIVATIVES. APPROXIMATION OF SECOND ORDER DERIVATIVES Second order dervatves appear n dffusve

More information

A Robust Method for Calculating the Correlation Coefficient

A Robust Method for Calculating the Correlation Coefficient A Robust Method for Calculatng the Correlaton Coeffcent E.B. Nven and C. V. Deutsch Relatonshps between prmary and secondary data are frequently quantfed usng the correlaton coeffcent; however, the tradtonal

More information

Application of B-Spline to Numerical Solution of a System of Singularly Perturbed Problems

Application of B-Spline to Numerical Solution of a System of Singularly Perturbed Problems Mathematca Aeterna, Vol. 1, 011, no. 06, 405 415 Applcaton of B-Splne to Numercal Soluton of a System of Sngularly Perturbed Problems Yogesh Gupta Department of Mathematcs Unted College of Engneerng &

More information

The Study of Teaching-learning-based Optimization Algorithm

The Study of Teaching-learning-based Optimization Algorithm Advanced Scence and Technology Letters Vol. (AST 06), pp.05- http://dx.do.org/0.57/astl.06. The Study of Teachng-learnng-based Optmzaton Algorthm u Sun, Yan fu, Lele Kong, Haolang Q,, Helongang Insttute

More information

12 MATH 101A: ALGEBRA I, PART C: MULTILINEAR ALGEBRA. 4. Tensor product

12 MATH 101A: ALGEBRA I, PART C: MULTILINEAR ALGEBRA. 4. Tensor product 12 MATH 101A: ALGEBRA I, PART C: MULTILINEAR ALGEBRA Here s an outlne of what I dd: (1) categorcal defnton (2) constructon (3) lst of basc propertes (4) dstrbutve property (5) rght exactness (6) localzaton

More information

Multigrid Methods and Applications in CFD

Multigrid Methods and Applications in CFD Multgrd Metods and Applcatons n CFD Mcael Wurst 0 May 009 Contents Introducton Typcal desgn of CFD solvers 3 Basc metods and ter propertes for solvng lnear systems of equatons 4 Geometrc Multgrd 3 5 Algebrac

More information

Additive Schwarz Method for DG Discretization of Anisotropic Elliptic Problems

Additive Schwarz Method for DG Discretization of Anisotropic Elliptic Problems Addtve Schwarz Method for DG Dscretzaton of Ansotropc Ellptc Problems Maksymlan Dryja 1, Potr Krzyżanowsk 1, and Marcus Sarks 2 1 Introducton In the paper we consder a second order ellptc problem wth dscontnuous

More information

Online Classification: Perceptron and Winnow

Online Classification: Perceptron and Winnow E0 370 Statstcal Learnng Theory Lecture 18 Nov 8, 011 Onlne Classfcaton: Perceptron and Wnnow Lecturer: Shvan Agarwal Scrbe: Shvan Agarwal 1 Introducton In ths lecture we wll start to study the onlne learnng

More information

Affine and Riemannian Connections

Affine and Riemannian Connections Affne and Remannan Connectons Semnar Remannan Geometry Summer Term 2015 Prof Dr Anna Wenhard and Dr Gye-Seon Lee Jakob Ullmann Notaton: X(M) space of smooth vector felds on M D(M) space of smooth functons

More information

Mean Field / Variational Approximations

Mean Field / Variational Approximations Mean Feld / Varatonal Appromatons resented by Jose Nuñez 0/24/05 Outlne Introducton Mean Feld Appromaton Structured Mean Feld Weghted Mean Feld Varatonal Methods Introducton roblem: We have dstrbuton but

More information

The Order Relation and Trace Inequalities for. Hermitian Operators

The Order Relation and Trace Inequalities for. Hermitian Operators Internatonal Mathematcal Forum, Vol 3, 08, no, 507-57 HIKARI Ltd, wwwm-hkarcom https://doorg/0988/mf088055 The Order Relaton and Trace Inequaltes for Hermtan Operators Y Huang School of Informaton Scence

More information

8.4 COMPLEX VECTOR SPACES AND INNER PRODUCTS

8.4 COMPLEX VECTOR SPACES AND INNER PRODUCTS SECTION 8.4 COMPLEX VECTOR SPACES AND INNER PRODUCTS 493 8.4 COMPLEX VECTOR SPACES AND INNER PRODUCTS All the vector spaces you have studed thus far n the text are real vector spaces because the scalars

More information

One-sided finite-difference approximations suitable for use with Richardson extrapolation

One-sided finite-difference approximations suitable for use with Richardson extrapolation Journal of Computatonal Physcs 219 (2006) 13 20 Short note One-sded fnte-dfference approxmatons sutable for use wth Rchardson extrapolaton Kumar Rahul, S.N. Bhattacharyya * Department of Mechancal Engneerng,

More information

New Method for Solving Poisson Equation. on Irregular Domains

New Method for Solving Poisson Equation. on Irregular Domains Appled Mathematcal Scences Vol. 6 01 no. 8 369 380 New Method for Solvng Posson Equaton on Irregular Domans J. Izadan and N. Karamooz Department of Mathematcs Facult of Scences Mashhad BranchIslamc Azad

More information

2.29 Numerical Fluid Mechanics Fall 2011 Lecture 12

2.29 Numerical Fluid Mechanics Fall 2011 Lecture 12 REVIEW Lecture 11: 2.29 Numercal Flud Mechancs Fall 2011 Lecture 12 End of (Lnear) Algebrac Systems Gradent Methods Krylov Subspace Methods Precondtonng of Ax=b FINITE DIFFERENCES Classfcaton of Partal

More information

Chapter 13: Multiple Regression

Chapter 13: Multiple Regression Chapter 13: Multple Regresson 13.1 Developng the multple-regresson Model The general model can be descrbed as: It smplfes for two ndependent varables: The sample ft parameter b 0, b 1, and b are used to

More information

Week 5: Neural Networks

Week 5: Neural Networks Week 5: Neural Networks Instructor: Sergey Levne Neural Networks Summary In the prevous lecture, we saw how we can construct neural networks by extendng logstc regresson. Neural networks consst of multple

More information

Lecture 5.8 Flux Vector Splitting

Lecture 5.8 Flux Vector Splitting Lecture 5.8 Flux Vector Splttng 1 Flux Vector Splttng The vector E n (5.7.) can be rewrtten as E = AU (5.8.1) (wth A as gven n (5.7.4) or (5.7.6) ) whenever, the equaton of state s of the separable form

More information

The Geometry of Logit and Probit

The Geometry of Logit and Probit The Geometry of Logt and Probt Ths short note s meant as a supplement to Chapters and 3 of Spatal Models of Parlamentary Votng and the notaton and reference to fgures n the text below s to those two chapters.

More information

Key Words: Hamiltonian systems, canonical integrators, symplectic integrators, Runge-Kutta-Nyström methods.

Key Words: Hamiltonian systems, canonical integrators, symplectic integrators, Runge-Kutta-Nyström methods. CANONICAL RUNGE-KUTTA-NYSTRÖM METHODS OF ORDERS 5 AND 6 DANIEL I. OKUNBOR AND ROBERT D. SKEEL DEPARTMENT OF COMPUTER SCIENCE UNIVERSITY OF ILLINOIS AT URBANA-CHAMPAIGN 304 W. SPRINGFIELD AVE. URBANA, ILLINOIS

More information

Foundations of Arithmetic

Foundations of Arithmetic Foundatons of Arthmetc Notaton We shall denote the sum and product of numbers n the usual notaton as a 2 + a 2 + a 3 + + a = a, a 1 a 2 a 3 a = a The notaton a b means a dvdes b,.e. ac = b where c s an

More information

The Exact Formulation of the Inverse of the Tridiagonal Matrix for Solving the 1D Poisson Equation with the Finite Difference Method

The Exact Formulation of the Inverse of the Tridiagonal Matrix for Solving the 1D Poisson Equation with the Finite Difference Method Journal of Electromagnetc Analyss and Applcatons, 04, 6, 0-08 Publshed Onlne September 04 n ScRes. http://www.scrp.org/journal/jemaa http://dx.do.org/0.46/jemaa.04.6000 The Exact Formulaton of the Inverse

More information

Module 9. Lecture 6. Duality in Assignment Problems

Module 9. Lecture 6. Duality in Assignment Problems Module 9 1 Lecture 6 Dualty n Assgnment Problems In ths lecture we attempt to answer few other mportant questons posed n earler lecture for (AP) and see how some of them can be explaned through the concept

More information

Chapter Newton s Method

Chapter Newton s Method Chapter 9. Newton s Method After readng ths chapter, you should be able to:. Understand how Newton s method s dfferent from the Golden Secton Search method. Understand how Newton s method works 3. Solve

More information

Structure and Drive Paul A. Jensen Copyright July 20, 2003

Structure and Drive Paul A. Jensen Copyright July 20, 2003 Structure and Drve Paul A. Jensen Copyrght July 20, 2003 A system s made up of several operatons wth flow passng between them. The structure of the system descrbes the flow paths from nputs to outputs.

More information

VQ widely used in coding speech, image, and video

VQ widely used in coding speech, image, and video at Scalar quantzers are specal cases of vector quantzers (VQ): they are constraned to look at one sample at a tme (memoryless) VQ does not have such constrant better RD perfomance expected Source codng

More information

Linear Approximation with Regularization and Moving Least Squares

Linear Approximation with Regularization and Moving Least Squares Lnear Approxmaton wth Regularzaton and Movng Least Squares Igor Grešovn May 007 Revson 4.6 (Revson : March 004). 5 4 3 0.5 3 3.5 4 Contents: Lnear Fttng...4. Weghted Least Squares n Functon Approxmaton...

More information

2.3 Nilpotent endomorphisms

2.3 Nilpotent endomorphisms s a block dagonal matrx, wth A Mat dm U (C) In fact, we can assume that B = B 1 B k, wth B an ordered bass of U, and that A = [f U ] B, where f U : U U s the restrcton of f to U 40 23 Nlpotent endomorphsms

More information

General viscosity iterative method for a sequence of quasi-nonexpansive mappings

General viscosity iterative method for a sequence of quasi-nonexpansive mappings Avalable onlne at www.tjnsa.com J. Nonlnear Sc. Appl. 9 (2016), 5672 5682 Research Artcle General vscosty teratve method for a sequence of quas-nonexpansve mappngs Cuje Zhang, Ynan Wang College of Scence,

More information

Convexity preserving interpolation by splines of arbitrary degree

Convexity preserving interpolation by splines of arbitrary degree Computer Scence Journal of Moldova, vol.18, no.1(52), 2010 Convexty preservng nterpolaton by splnes of arbtrary degree Igor Verlan Abstract In the present paper an algorthm of C 2 nterpolaton of dscrete

More information

Kernel Methods and SVMs Extension

Kernel Methods and SVMs Extension Kernel Methods and SVMs Extenson The purpose of ths document s to revew materal covered n Machne Learnng 1 Supervsed Learnng regardng support vector machnes (SVMs). Ths document also provdes a general

More information

The interface control domain decomposition (ICDD) method for the Stokes problem. (Received: 15 July Accepted: 13 September 2013)

The interface control domain decomposition (ICDD) method for the Stokes problem. (Received: 15 July Accepted: 13 September 2013) Journal of Coupled Systems Multscale Dynamcs Copyrght 2013 by Amercan Scentfc Publshers All rghts reserved. Prnted n the Unted States of Amerca do:10.1166/jcsmd.2013.1026 J. Coupled Syst. Multscale Dyn.

More information

where the sums are over the partcle labels. In general H = p2 2m + V s(r ) V j = V nt (jr, r j j) (5) where V s s the sngle-partcle potental and V nt

where the sums are over the partcle labels. In general H = p2 2m + V s(r ) V j = V nt (jr, r j j) (5) where V s s the sngle-partcle potental and V nt Physcs 543 Quantum Mechancs II Fall 998 Hartree-Fock and the Self-consstent Feld Varatonal Methods In the dscusson of statonary perturbaton theory, I mentoned brey the dea of varatonal approxmaton schemes.

More information

CSci 6974 and ECSE 6966 Math. Tech. for Vision, Graphics and Robotics Lecture 21, April 17, 2006 Estimating A Plane Homography

CSci 6974 and ECSE 6966 Math. Tech. for Vision, Graphics and Robotics Lecture 21, April 17, 2006 Estimating A Plane Homography CSc 6974 and ECSE 6966 Math. Tech. for Vson, Graphcs and Robotcs Lecture 21, Aprl 17, 2006 Estmatng A Plane Homography Overvew We contnue wth a dscusson of the major ssues, usng estmaton of plane projectve

More information

COMPARISON OF SOME RELIABILITY CHARACTERISTICS BETWEEN REDUNDANT SYSTEMS REQUIRING SUPPORTING UNITS FOR THEIR OPERATIONS

COMPARISON OF SOME RELIABILITY CHARACTERISTICS BETWEEN REDUNDANT SYSTEMS REQUIRING SUPPORTING UNITS FOR THEIR OPERATIONS Avalable onlne at http://sck.org J. Math. Comput. Sc. 3 (3), No., 6-3 ISSN: 97-537 COMPARISON OF SOME RELIABILITY CHARACTERISTICS BETWEEN REDUNDANT SYSTEMS REQUIRING SUPPORTING UNITS FOR THEIR OPERATIONS

More information

Non shape regular domain decompositions: an analysis using a stable decomposition in H 1 0

Non shape regular domain decompositions: an analysis using a stable decomposition in H 1 0 on shape regular doman decompostons: an analyss usng a stable decomposton n 1 0 Martn J. Gander 1, Laurence alpern, and Kévn Santugn Repquet 3 1 Unversté de Genève, Secton de Mathématques, Martn.Gander@unge.ch

More information

ON A DETERMINATION OF THE INITIAL FUNCTIONS FROM THE OBSERVED VALUES OF THE BOUNDARY FUNCTIONS FOR THE SECOND-ORDER HYPERBOLIC EQUATION

ON A DETERMINATION OF THE INITIAL FUNCTIONS FROM THE OBSERVED VALUES OF THE BOUNDARY FUNCTIONS FOR THE SECOND-ORDER HYPERBOLIC EQUATION Advanced Mathematcal Models & Applcatons Vol.3, No.3, 2018, pp.215-222 ON A DETERMINATION OF THE INITIAL FUNCTIONS FROM THE OBSERVED VALUES OF THE BOUNDARY FUNCTIONS FOR THE SECOND-ORDER HYPERBOLIC EUATION

More information

DUE: WEDS FEB 21ST 2018

DUE: WEDS FEB 21ST 2018 HOMEWORK # 1: FINITE DIFFERENCES IN ONE DIMENSION DUE: WEDS FEB 21ST 2018 1. Theory Beam bendng s a classcal engneerng analyss. The tradtonal soluton technque makes smplfyng assumptons such as a constant

More information

Chapter 12. Ordinary Differential Equation Boundary Value (BV) Problems

Chapter 12. Ordinary Differential Equation Boundary Value (BV) Problems Chapter. Ordnar Dfferental Equaton Boundar Value (BV) Problems In ths chapter we wll learn how to solve ODE boundar value problem. BV ODE s usuall gven wth x beng the ndependent space varable. p( x) q(

More information

THE SUMMATION NOTATION Ʃ

THE SUMMATION NOTATION Ʃ Sngle Subscrpt otaton THE SUMMATIO OTATIO Ʃ Most of the calculatons we perform n statstcs are repettve operatons on lsts of numbers. For example, we compute the sum of a set of numbers, or the sum of the

More information

Solving Nonlinear Differential Equations by a Neural Network Method

Solving Nonlinear Differential Equations by a Neural Network Method Solvng Nonlnear Dfferental Equatons by a Neural Network Method Luce P. Aarts and Peter Van der Veer Delft Unversty of Technology, Faculty of Cvlengneerng and Geoscences, Secton of Cvlengneerng Informatcs,

More information

CHAPTER 5: Lie Differentiation and Angular Momentum

CHAPTER 5: Lie Differentiation and Angular Momentum CHAPTER 5: Le Dfferentaton and Angular Momentum Jose G. Vargas 1 Le dfferentaton Kähler s theory of angular momentum s a specalzaton of hs approach to Le dfferentaton. We could deal wth the former drectly,

More information

36.1 Why is it important to be able to find roots to systems of equations? Up to this point, we have discussed how to find the solution to

36.1 Why is it important to be able to find roots to systems of equations? Up to this point, we have discussed how to find the solution to ChE Lecture Notes - D. Keer, 5/9/98 Lecture 6,7,8 - Rootndng n systems o equatons (A) Theory (B) Problems (C) MATLAB Applcatons Tet: Supplementary notes rom Instructor 6. Why s t mportant to be able to

More information

Advanced Circuits Topics - Part 1 by Dr. Colton (Fall 2017)

Advanced Circuits Topics - Part 1 by Dr. Colton (Fall 2017) Advanced rcuts Topcs - Part by Dr. olton (Fall 07) Part : Some thngs you should already know from Physcs 0 and 45 These are all thngs that you should have learned n Physcs 0 and/or 45. Ths secton s organzed

More information