Approximate Dynamic Programming by Linear Programming for Stochastic Scheduling

Size: px
Start display at page:

Download "Approximate Dynamic Programming by Linear Programming for Stochastic Scheduling"

Transcription

1 Approximate Dyamic Programmig by Liear Programmig for Stochastic Schedulig Mohamed Mostagir Nelso Uha 1 Itroductio I stochastic schedulig, we wat to allocate a limited amout of resources to a set of jobs that eed to be serviced. Ulike i determiistic schedulig, however, the parameters of the system may be stochastic. For example, the time it takes to process a job may be subject to radom fluctuatios. Stochastic schedulig problems occur i a variety of practical situatios, such as maufacturig, costructio, ad compiler optimizatio. As i determiistic schedulig, the set of stochastic schedulig problems is icredibly large ad diverse. Oe importat class of models ivolves schedulig a fixed umber of jobs o a fixed umber of idetical parallel machies while miimizig a give performace measure. The processig times of the jobs are assumed to follow some joit probability distributio. I additio, there may be precedece costraits, or iterdepedecies betwee the jobs that require certai jobs ot be scheduled util others are completed. The determiistic couterpart to this class of problems has bee studied extesively [KSW98]. A aïve approach to these problems would be to take the expected processig times ad use the algorithms for the determiistic problems. Ufortuately, it is easy to costruct examples that shows that this approach ca produce solutios that are arbitrarily bad. Fortuately, however, this class of stochastic schedulig problems ca be easily cast as a Markov decisio process (MDP) ad therefore ca be attacked by dyamic programmig methods. Results for this class of stochastic schedulig problems are somewhat scattered, ad have bee obtaied usig a variety of methods. Rothkopf [R66] showed that for oe machie without precedece costraits, a idex rule miimizes the expected sum of weighted completio times for arbitrary processig time probability distributios. Möhrig, Radermacher ad Weiss [MRW84, MRW85] study the aalytic properties of various classes of schedulig policies, ad determie optimal policies for special cases. Möhrig, Schulz, ad Uetz [MSU99] developed approximatio algorithms for a variety of stochastic 1

2 schedulig problems usig techiques from combiatorial optimizatio. Bertsekas ad Castaño [BC99] focus o a differet class of stochastic schedulig problems, the quiz problem ad its variatios. They show how rollout algorithms ca be implemeted efficietly, ad preset experimetal evidece that the performace of rollout policies is ear-optimal. For this project, we cosider a problem with oe machie ad a arbitrary ormalized regular ad additive objective fuctio. We recast our fiite-horizo decisio problem ito a stochastic shortest path problem. We show that for a relaxed formulatio of our problem, the error of the solutio obtaied by the approximate liear programmig approach to dyamic programmig as preseted i de Farias ad Va Roy [dv03] is uiformly bouded over the umber of jobs that eed to be scheduled, provided that the expected job processig times are fiite. Fially, we argue usig results from dyamic programmig that the approximate solutio for the relaxed formulatio of our problem is also ot that far away from the optimal solutio of the origial problem. 2 The problem Cosider the followig problem. We have a set of jobs N = {1,...,} to be processed o oe machie. The machie ca oly process oe job at a time. The processig time of job i follows a discrete probability distributio p i, ad the distributios p 1,..., p are assumed to be pairwise stochastically idepedet. The jobs have to be scheduled opreemptively, that is, oce a job has started, it must be processed cotiuously util it is fiished. The schedule must also respect ay precedece costraits, or iterdepedecies betwee jobs that require certai jobs ot be scheduled before others are completed. We would like to miimize the expected value of the objective fuctio ( ) 1 ( ) γ C (1),...,C () = 1 h (R i ) C (i+1) C (i) i=0 where C (i) deotes the time of the ith job completio (C (0) = 0), R i is the set of jobs remaiig to be processed at the time of the ith job completio, ad h is a set fuctio such that h ( ) = 0. Such a objective fuctio is said to be additive. The fuctio h ca be iterpreted as the holdig cost per uit time o the set of ucompleted jobs. We also assume that γ is odecreasig i the job completio times. May commo objective fuctios i schedulig are odecreasig ad additive. For example, h (S) = j S w j for all S N geerates the ormalized 1 sum of weighted completio times, j N w j C j. 2.1 Formulatio as a fiite horizo MDP We ca formulate this problem as a fiite horizo MDP with a fiite state space S. For each state x S, there is a set of available actios A x. Takig actio 2

3 a A x i state x, we trasitio to state y with probability P a (x, y) ad icur cost g a (x, y). Sice we oly have oe machie, the state of the system x is sufficietly characterized by the set of jobs remaiig to be processed R x ad the maximum completio time of the jobs completed so far, C max (x): x = (C max (x),r x ) S. The size of the state space is expoetial i the umber of jobs. Sice the objective fuctio is odecreasig i completio times, ad sice the jobs processig times are idepedet of their start times, ay optimal policy eed ot leave the machie deliberately idle at ay time. Therefore, a actio i our problem is simply the ext job to be processed: at every state x, the actio set is A x R x. If there are o precedece costraits, A x = R x. The time stage costs are 1 g a (x, y) = h (R x )(C max (y) C max (x)). The trasitio probabilities are { p a (t) if R y = R x \ {a} ad C max (y) = C max (x) + t P a (x, y) = 0 otherwise. The problem is the to solve for the followig fiite-horizo cost-to-go fuctio J (x, ) = 0 J (x, t) = mi P a (x, y)(g a (x, y) + J (y, t + 1)), t = 0, 1,..., 1. a A x 2.2 Reformulatio ito a stochastic shortest path problem Sice our state space is expoetial i size, we caot hope to solve our problem exactly usig dyamic programmig methods. All of the major methods for approximate dyamic programmig cosider ifiite-horizo discouted cost problems. I order to use these methods for our fiite-horizo stochastic schedulig problem, we recast our problem ito a stochastic shortest path (SSP) problem. We refer to the followig formulatio of our problem as the origial SSP formulatio. We itroduce a termiatig state x with the followig property: oly the states that have o more remaiig jobs (R x = ) ca reach x i oe step. Observe that for all states such that R x = ad at the termiatig state x, the set of actios available A x is empty, ad therefore the trasitio probabilities ad time stage costs ivolvig x are ot affected by what actio is take. The trasitio probabilities ivolvig x are 1 x such that R x = P a (x, x) = 1 if x = x 0 otherwise 3

4 ad the time stage costs ivolvig x are { 0 x such that R x = g a (x, x) = 0 if x = x. The cost-to-go fuctio is therefore T(x) J (x) = mi E g u (x t, x t+1 ) x 0 = x. u where T (x) is the time stage whe the system reaches the termiatig state. Recall that a statioary policy u is called a proper policy if, whe usig this policy, there is positive probability that the termiatig state will be reached after a fiite umber of stages. Also recall that if all statioary policies are proper, the the cost-to-go fuctio for the SSP problem is the uique solutio to Bellma s equatio [B01]. Sice ay policy i our schedulig problem requires oe additioal job to be scheduled at each time stage, we must always reach the termiatig state with probability 1, regardless of the policy used. Therefore, to solve our problem exactly, we ca solve Bellma s equatio. 3 Solutio method ad bouds We cosider the liear programmig approach to dyamic programmig by de Farias ad Va Roy [dv03] as a solutio method to our problem. We briefly review the method ad related results. 3.1 The liear programmig approach to dyamic programmig Defie the operator TJ = mi {g u + αp u J}. u where the miimizatio is carried out compoet-wise. We wat to determie,α J, the uique solutio to Bellma s equatio, TJ = J. The exact liear programmig (ELP) approach to solvig Bellma s equatio solves the followig optimizatio problem for ay c > 0: maximize T c J subject to TJ J. Note that this optimizatio problem ca be recast as the followig liear program: T maximize c J subject to g a (x) + α P a (x, y)j (y) J (y), x S, a A x. 4

5 The approximate liear programmig (ALP) approach to dyamic programmig reduces the umber of variables i the exact liear programmig by the use of basis fuctios. Give a set of basis fuctios φ 1,...,φ K, we defie the matrix Φ = φ 1 φ K, ad i the hopes of computig a vector r such that Φ r is a close approximatio,α to J, we solve the followig optimizatio problem: maximize subject to T c Φr (1) TΦr Φr. This problem ca be recast as a liear program just as i the ELP approach. Give a solutio r, we hopefully ca obtai a good policy by usig the greedy policy with respect to Φ r, u (x) = arg mi g a (x) + α P a (x, y)(φ r)(y). a A x de Farias ad Va Roy [dv03] proved the followig boud o the error of the ALP solutio. Theorem 3.1. (de Farias ad Va Roy 2003) Let r be a solutio of the approximate LP (1), ad α (0, 1). The, for ay v R K such that (Φv) (x) > 0 for all x S ad αhφv < Φv,,α J 2c T Φv Φ r 1,c mi J 1 β Φv r where (HΦv)(x) = max P a (x, y)(φv) (y) a A x ad,α α (HΦv)(x) β Φv = max. x (Φv) (x) Φr 3.2 A uiform boud over umber of jobs,1/φv We relax our stochastic shortest path problem by itroducig discout factors α (0, 1) at each time stage. We refer to this formulatio as the α-relaxed SSP formulatio. The cost-to-go fuctio for this problem is T(x),α J (x) = mi E α t g u (x t, x t+1 ) x 0 = x. u For this relaxatio, we show that the error of the approximate liear programmig solutio is uiformly bouded over the umber of jobs to be scheduled. 5

6 Theorem 3.2. Assume that the holdig cost h (S) is bouded above by M for all subsets S of N. Let r be the ALP solutio to the α-relaxed SSP formulatio of the stochastic schedulig problem. For α (0, 1),,α 2M max i N E [p i ] J Φ r 1,c. Proof. Due to the ature of our problem, we kow that for ay state x, T (x) = R x +1. However, sice g a (x, x) = 0 for all x such that R x =, ad g a ( x, x) = 0, whe computig the cost-to-go fuctio at state x, we oly eed to cosider time stage costs for time stages 0 through R x 1. Sice the holdig cost h is bouded by above by M for all subsets of N, for some policy u we have R x 1,α J (x) = mi E α t g u(xt) (x t, x t+1 ) x 0 = x u R x 1 E α t g u(xt) (x t, x t+1 ) x 0 = x R x 1 E g u(xt) (x t, x t+1 ) x 0 = x R x 1 1 = E h (R xt )(C max (x t+1 ) C max (x t )) x0 = x, u R x 1 1 = E h (R xt )p u(xt) x 0 = x R x 1 1 E Mp u(xt) x 0 = x M = E [p i ] where p u(xt) is the processig time of the job chose at state x t. Cosider the fuctio k V (x) = E [p i ] for some costat k. The for all x S α (HV )(x) = α max P a (x, y)v (y) a A x 1 = αk max P a (x, y) E [p i ] a A x i R y 6

7 ( ) k k = α E [p i ] E [p a ] ( ) αe [p a ] = V (x) α E [p i ] < V (x) where a arg max a Ax P a (x, y)v (y). Therefore, V is a Lyapuov fuctio, ad there is a β < 1 idepedet of the umber of jobs such that αhv βv. Also ote that,α mi J r Φr J,α,1/V,1/V = max x S max,α (x) J V (x) [ ] M E x S k M = k [ i Rx p ] p i This is uiformly bouded over the umber of states. Fially, we cosider c T V. Let c be some probability distributio such that c (x) > 0 for all x S. ( ) k c (x) V (x) = c (x) E [p i ] x S x S ( ) 1 = k c (x) E [p i ] x S ( ) k c (x) max E [p i ] i N x S = k max E [p i ]. i N This is uiformly bouded over the umber of jobs, provided that the expected processig times of the jobs are bouded. Therefore, by Theorem 3.1, provided that E [p i ] is oe of our basis fuctios, we have that,α J 2c T Φv Φ r 1,c mi J α Φr 1,1/Φv βφv r 2M max i N E [p i ] which is uiformly bouded over the umber of jobs,. i 7

8 3.3 What happes whe α = 1? Ufortuately, the boud i Theorem 3.2 explodes for α = 1, ad therefore does ot directly apply to the origial SSP formulatio. The problem lies i the choice of the Lyapuov fuctio. If we use the same Lyapuov fuctio as i the proof of Theorem 3.2, we are left with the followig equality ( ) E [p a ] (HV )(x) = V (x) 1. E [p i ] As the umber of jobs, the ratio of the expected processig time of ay oe job to the sum of the expected processig times of all jobs goes to zero. Therefore, by usig the methods from the proof of Theorem 3.2, we caot fid a β < 1 such that HV βv regardless of the umber of jobs. However, sice the ALP solutio for the α-relaxed formulatio is ot that far away from the optimal solutio to the α-relaxed formulatio, we ca show that the ALP solutio obtaied for the α-relaxed SSP formulatio is also ot that far away from the optimal solutio of the origial SSP formulatio, depedig o the value of α. First, we briefly preset a well-kow result from dyamic programmig that relates the costs of the ifiite-horizo discouted cost ad average-cost problems. Lemma 3.1. For ay statioary policy u ad α (0, 1), we have J u J α,u = + h u + O ( ) where ( ) N 1 J α,u = α k k 1 P u g u, J u = lim Pu k g u, N N k=0 k=0 h u is a vector satisfyig J u + h u = g u + P u h u, ad O ( ) is a α-depedet vector such that lim O ( ) = 0. α 1 For the stochastic schedulig problem formulated as a stochastic shortest path problem, the J u i Lemma 3.1 is equal to zero. Recall that J is the cost-to-go fuctio for the origial SSP formulatio. Therefore, Lemma 3.1 implies,α,α J J J J = O ( ) J J,α O ( ) where J is the cost of usig the optimal greedy policy for the α-relaxed SSP formulatio i the origial SSP formulatio (h u i Lemma 3.1). Sice the costto-go fuctios for the origial ad α-relaxed SSP formulatios are ot that far 8

9 apart whe α is large, the ALP solutio for the α-relaxed SSP formulatio is ot that far away from the optimal solutio to the origial SSP formulatio whe α is large. Corollary 3.1. Let r be the optimal solutio to the ALP for the α-relaxed SSP formulatio. The for α (0, 1), J Φ r 1,c 2M max i N E [p i ] T c O ( ). Proof. The result follows immediately from Theorem 3.2 ad the argumets above: J Φ r J J,α,α 1,c 1,c + J Φ r 1,c T c J J,α + 2H max i N E [p i ] 2H max i N E [p i ] T c O ( ) Although we have had difficulty i obtaiig a error boud uiform over the umber of jobs for the ALP solutio whe α = 1, recet work [d04] idicates that relaxig the problem to iclude discout factors is the right approach to obtaiig good bouds for the origial SSP formulatio. 4 Coclusio We have show that i theory, the approximate liear programmig approach to dyamic programmig should be a good solutio method for the stochastic schedulig problem we studied. By relaxig the problem to iclude a discoutig factor at each time stage, we obtaied a boud o the error of the ALP solutio that is uiform over the size of the problem. Usig well-kow results i dyamic programmig, we also showed that by solvig the relaxed problem, we obtai a solutio that is ot that far away from the optimal solutio to the origial problem. The aalysis above also provides a error boud uiform over the umber of jobs for the multiple machie case. Although the state ad actio spaces of the problem with multiple machies are very complex, a upper boud o the cost-to-go fuctio at ay state ca always be obtaied by cosiderig the cost-to-go fuctio o just oe machie. Oe immediate aveue of further research would be to determie a error boud that is uiform over the umber of jobs for the ALP solutio to the origial SSP formulatio. Although our error boud does ot grow as the size of the problem grows, the error boud is still potetially very large. It would be ice to see if a tighter boud o the errors ca be obtaied. Aother directio of ivestigatio would be to ru computatioal experimets o usig ALP 9

10 for our stochastic schedulig problem. It would be ice to see how the ALP approach performs i practice, ad whether or ot i reality ALP solves the origial SSP formulatio (with α = 1) poorly as the size of the problem grows. Fially, the ALP method relies o solvig a liear program with a costrait for every state-actio pair, which i our problem would result i a extremely large optimizatio problem. Oe might ivestigate how the costrait samplig LP approach to dyamic programmig [dv01] performs with the stochastic schedulig problem we studied, both theoretically ad practically. Refereces [B01] [BC99] Bertsekas, D. (2001). Dyamic Programmig ad Optimal Cotrol. Athea Scietific, Belmot MA. Bertsekas, D., D. Castaño (1999). Rollout algorithms for stochastic schedulig problems. Joural of Heuristics 5, pp [d04] de Farias, D. P. (2004). Private commuicatio, May 10, [dv03] [dv01] de Farias, D. P., B. Va Roy (2003). The liear programmig approach to approximate dyamic programmig. Operatios Research 51, pp de Farias, D. P., B. Va Roy (2001). O costrait samplig for the liear programmig approach to dyamic programmig. Mathematics of Operatios Research, forthcomig. [KSW98] Karger, D., C. Stei, J. Wei (1998). Schedulig algorithms. CRC Algorithms ad Theory of Computatio Hadbook, Chapter 35. [MRW84] Möhrig, R. H., F. J. Radermacher, G. Weiss (1984). Stochastic schedulig problems I: geeral strategies. Zeitschrift für Operatios Research 28, pp [MRW85] Möhrig, R. H., F. J. Radermacher, G. Weiss (1985). Stochastic schedulig problems II: set strategies. Zeitschrift für Operatios Research 29, pp [MSU99] Möhrig, R. H., A. S. Schulz, M. Uetz (1999). Approximatio i stochastic schedulig: the power of LP-based priority policies. Joural of the ACM 46, pp [R66] [U96] Rothkopf, M. H. (1966). Schedulig with radom service times. Maagemet Sciece 12, pp Uetz, M. (1996). Algorithms for determiistic ad stochastic schedulig. Ph.D. dissertatio, Istitut für Mathematik, Techische Uiversität Berli, Germay. 10

Markov Decision Processes

Markov Decision Processes Markov Decisio Processes Defiitios; Statioary policies; Value improvemet algorithm, Policy improvemet algorithm, ad liear programmig for discouted cost ad average cost criteria. Markov Decisio Processes

More information

w (1) ˆx w (1) x (1) /ρ and w (2) ˆx w (2) x (2) /ρ.

w (1) ˆx w (1) x (1) /ρ and w (2) ˆx w (2) x (2) /ρ. 2 5. Weighted umber of late jobs 5.1. Release dates ad due dates: maximimizig the weight of o-time jobs Oce we add release dates, miimizig the umber of late jobs becomes a sigificatly harder problem. For

More information

Lecture 7: October 18, 2017

Lecture 7: October 18, 2017 Iformatio ad Codig Theory Autum 207 Lecturer: Madhur Tulsiai Lecture 7: October 8, 207 Biary hypothesis testig I this lecture, we apply the tools developed i the past few lectures to uderstad the problem

More information

ACO Comprehensive Exam 9 October 2007 Student code A. 1. Graph Theory

ACO Comprehensive Exam 9 October 2007 Student code A. 1. Graph Theory 1. Graph Theory Prove that there exist o simple plaar triagulatio T ad two distict adjacet vertices x, y V (T ) such that x ad y are the oly vertices of T of odd degree. Do ot use the Four-Color Theorem.

More information

Optimization Methods MIT 2.098/6.255/ Final exam

Optimization Methods MIT 2.098/6.255/ Final exam Optimizatio Methods MIT 2.098/6.255/15.093 Fial exam Date Give: December 19th, 2006 P1. [30 pts] Classify the followig statemets as true or false. All aswers must be well-justified, either through a short

More information

A New Solution Method for the Finite-Horizon Discrete-Time EOQ Problem

A New Solution Method for the Finite-Horizon Discrete-Time EOQ Problem This is the Pre-Published Versio. A New Solutio Method for the Fiite-Horizo Discrete-Time EOQ Problem Chug-Lu Li Departmet of Logistics The Hog Kog Polytechic Uiversity Hug Hom, Kowloo, Hog Kog Phoe: +852-2766-7410

More information

Definitions and Theorems. where x are the decision variables. c, b, and a are constant coefficients.

Definitions and Theorems. where x are the decision variables. c, b, and a are constant coefficients. Defiitios ad Theorems Remember the scalar form of the liear programmig problem, Miimize, Subject to, f(x) = c i x i a 1i x i = b 1 a mi x i = b m x i 0 i = 1,2,, where x are the decisio variables. c, b,

More information

Machine Learning Brett Bernstein

Machine Learning Brett Bernstein Machie Learig Brett Berstei Week 2 Lecture: Cocept Check Exercises Starred problems are optioal. Excess Risk Decompositio 1. Let X = Y = {1, 2,..., 10}, A = {1,..., 10, 11} ad suppose the data distributio

More information

Information-based Feature Selection

Information-based Feature Selection Iformatio-based Feature Selectio Farza Faria, Abbas Kazeroui, Afshi Babveyh Email: {faria,abbask,afshib}@staford.edu 1 Itroductio Feature selectio is a topic of great iterest i applicatios dealig with

More information

IP Reference guide for integer programming formulations.

IP Reference guide for integer programming formulations. IP Referece guide for iteger programmig formulatios. by James B. Orli for 15.053 ad 15.058 This documet is iteded as a compact (or relatively compact) guide to the formulatio of iteger programs. For more

More information

Stochastic Matrices in a Finite Field

Stochastic Matrices in a Finite Field Stochastic Matrices i a Fiite Field Abstract: I this project we will explore the properties of stochastic matrices i both the real ad the fiite fields. We first explore what properties 2 2 stochastic matrices

More information

REGRESSION WITH QUADRATIC LOSS

REGRESSION WITH QUADRATIC LOSS REGRESSION WITH QUADRATIC LOSS MAXIM RAGINSKY Regressio with quadratic loss is aother basic problem studied i statistical learig theory. We have a radom couple Z = X, Y ), where, as before, X is a R d

More information

Expectation and Variance of a random variable

Expectation and Variance of a random variable Chapter 11 Expectatio ad Variace of a radom variable The aim of this lecture is to defie ad itroduce mathematical Expectatio ad variace of a fuctio of discrete & cotiuous radom variables ad the distributio

More information

An Introduction to Randomized Algorithms

An Introduction to Randomized Algorithms A Itroductio to Radomized Algorithms The focus of this lecture is to study a radomized algorithm for quick sort, aalyze it usig probabilistic recurrece relatios, ad also provide more geeral tools for aalysis

More information

Basics of Probability Theory (for Theory of Computation courses)

Basics of Probability Theory (for Theory of Computation courses) Basics of Probability Theory (for Theory of Computatio courses) Oded Goldreich Departmet of Computer Sciece Weizma Istitute of Sciece Rehovot, Israel. oded.goldreich@weizma.ac.il November 24, 2008 Preface.

More information

9 - Markov processes and Burt & Allison 1963 AGEC

9 - Markov processes and Burt & Allison 1963 AGEC This documet was geerated at 4:51 PM o Wedesday, October 09, 2013 Copyright 2013 Richard T. Woodward 9 - Markov processes ad Burt & Alliso 1963 AGEC 637-2013 I. What is a Markov Chai? A Markov chai is

More information

Optimally Sparse SVMs

Optimally Sparse SVMs A. Proof of Lemma 3. We here prove a lower boud o the umber of support vectors to achieve geeralizatio bouds of the form which we cosider. Importatly, this result holds ot oly for liear classifiers, but

More information

Queuing Theory. Basic properties, Markovian models, Networks of queues, General service time distributions, Finite source models, Multiserver queues

Queuing Theory. Basic properties, Markovian models, Networks of queues, General service time distributions, Finite source models, Multiserver queues Queuig Theory Basic properties, Markovia models, Networks of queues, Geeral service time distributios, Fiite source models, Multiserver queues Chapter 8 Kedall s Notatio for Queuig Systems A/B/X/Y/Z: A

More information

c. Explain the basic Newsvendor model. Why is it useful for SC models? e. What additional research do you believe will be helpful in this area?

c. Explain the basic Newsvendor model. Why is it useful for SC models? e. What additional research do you believe will be helpful in this area? 1. Research Methodology a. What is meat by the supply chai (SC) coordiatio problem ad does it apply to all types of SC s? Does the Bullwhip effect relate to all types of SC s? Also does it relate to SC

More information

Stochastic Simulation

Stochastic Simulation Stochastic Simulatio 1 Itroductio Readig Assigmet: Read Chapter 1 of text. We shall itroduce may of the key issues to be discussed i this course via a couple of model problems. Model Problem 1 (Jackso

More information

TEACHER CERTIFICATION STUDY GUIDE

TEACHER CERTIFICATION STUDY GUIDE COMPETENCY 1. ALGEBRA SKILL 1.1 1.1a. ALGEBRAIC STRUCTURES Kow why the real ad complex umbers are each a field, ad that particular rigs are ot fields (e.g., itegers, polyomial rigs, matrix rigs) Algebra

More information

Lecture 19: Convergence

Lecture 19: Convergence Lecture 19: Covergece Asymptotic approach I statistical aalysis or iferece, a key to the success of fidig a good procedure is beig able to fid some momets ad/or distributios of various statistics. I may

More information

Uniform Strict Practical Stability Criteria for Impulsive Functional Differential Equations

Uniform Strict Practical Stability Criteria for Impulsive Functional Differential Equations Global Joural of Sciece Frotier Research Mathematics ad Decisio Scieces Volume 3 Issue Versio 0 Year 03 Type : Double Blid Peer Reviewed Iteratioal Research Joural Publisher: Global Jourals Ic (USA Olie

More information

Random assignment with integer costs

Random assignment with integer costs Radom assigmet with iteger costs Robert Parviaie Departmet of Mathematics, Uppsala Uiversity P.O. Box 480, SE-7506 Uppsala, Swede robert.parviaie@math.uu.se Jue 4, 200 Abstract The radom assigmet problem

More information

THE ASYMPTOTIC COMPLEXITY OF MATRIX REDUCTION OVER FINITE FIELDS

THE ASYMPTOTIC COMPLEXITY OF MATRIX REDUCTION OVER FINITE FIELDS THE ASYMPTOTIC COMPLEXITY OF MATRIX REDUCTION OVER FINITE FIELDS DEMETRES CHRISTOFIDES Abstract. Cosider a ivertible matrix over some field. The Gauss-Jorda elimiatio reduces this matrix to the idetity

More information

Lecture 6 Chi Square Distribution (χ 2 ) and Least Squares Fitting

Lecture 6 Chi Square Distribution (χ 2 ) and Least Squares Fitting Lecture 6 Chi Square Distributio (χ ) ad Least Squares Fittig Chi Square Distributio (χ ) Suppose: We have a set of measuremets {x 1, x, x }. We kow the true value of each x i (x t1, x t, x t ). We would

More information

Generalized Semi- Markov Processes (GSMP)

Generalized Semi- Markov Processes (GSMP) Geeralized Semi- Markov Processes (GSMP) Summary Some Defiitios Markov ad Semi-Markov Processes The Poisso Process Properties of the Poisso Process Iterarrival times Memoryless property ad the residual

More information

There is no straightforward approach for choosing the warmup period l.

There is no straightforward approach for choosing the warmup period l. B. Maddah INDE 504 Discrete-Evet Simulatio Output Aalysis () Statistical Aalysis for Steady-State Parameters I a otermiatig simulatio, the iterest is i estimatig the log ru steady state measures of performace.

More information

6.3 Testing Series With Positive Terms

6.3 Testing Series With Positive Terms 6.3. TESTING SERIES WITH POSITIVE TERMS 307 6.3 Testig Series With Positive Terms 6.3. Review of what is kow up to ow I theory, testig a series a i for covergece amouts to fidig the i= sequece of partial

More information

ECE-S352 Introduction to Digital Signal Processing Lecture 3A Direct Solution of Difference Equations

ECE-S352 Introduction to Digital Signal Processing Lecture 3A Direct Solution of Difference Equations ECE-S352 Itroductio to Digital Sigal Processig Lecture 3A Direct Solutio of Differece Equatios Discrete Time Systems Described by Differece Equatios Uit impulse (sample) respose h() of a DT system allows

More information

Lecture 6 Chi Square Distribution (χ 2 ) and Least Squares Fitting

Lecture 6 Chi Square Distribution (χ 2 ) and Least Squares Fitting Lecture 6 Chi Square Distributio (χ ) ad Least Squares Fittig Chi Square Distributio (χ ) Suppose: We have a set of measuremets {x 1, x, x }. We kow the true value of each x i (x t1, x t, x t ). We would

More information

Optimization Methods: Linear Programming Applications Assignment Problem 1. Module 4 Lecture Notes 3. Assignment Problem

Optimization Methods: Linear Programming Applications Assignment Problem 1. Module 4 Lecture Notes 3. Assignment Problem Optimizatio Methods: Liear Programmig Applicatios Assigmet Problem Itroductio Module 4 Lecture Notes 3 Assigmet Problem I the previous lecture, we discussed about oe of the bech mark problems called trasportatio

More information

62. Power series Definition 16. (Power series) Given a sequence {c n }, the series. c n x n = c 0 + c 1 x + c 2 x 2 + c 3 x 3 +

62. Power series Definition 16. (Power series) Given a sequence {c n }, the series. c n x n = c 0 + c 1 x + c 2 x 2 + c 3 x 3 + 62. Power series Defiitio 16. (Power series) Give a sequece {c }, the series c x = c 0 + c 1 x + c 2 x 2 + c 3 x 3 + is called a power series i the variable x. The umbers c are called the coefficiets of

More information

Analysis of the Chow-Robbins Game with Biased Coins

Analysis of the Chow-Robbins Game with Biased Coins Aalysis of the Chow-Robbis Game with Biased Cois Arju Mithal May 7, 208 Cotets Itroductio to Chow-Robbis 2 2 Recursive Framework for Chow-Robbis 2 3 Geeralizig the Lower Boud 3 4 Geeralizig the Upper Boud

More information

Research Article A Unified Weight Formula for Calculating the Sample Variance from Weighted Successive Differences

Research Article A Unified Weight Formula for Calculating the Sample Variance from Weighted Successive Differences Discrete Dyamics i Nature ad Society Article ID 210761 4 pages http://dxdoiorg/101155/2014/210761 Research Article A Uified Weight Formula for Calculatig the Sample Variace from Weighted Successive Differeces

More information

Discrete-Time Systems, LTI Systems, and Discrete-Time Convolution

Discrete-Time Systems, LTI Systems, and Discrete-Time Convolution EEL5: Discrete-Time Sigals ad Systems. Itroductio I this set of otes, we begi our mathematical treatmet of discrete-time s. As show i Figure, a discrete-time operates or trasforms some iput sequece x [

More information

Properties and Hypothesis Testing

Properties and Hypothesis Testing Chapter 3 Properties ad Hypothesis Testig 3.1 Types of data The regressio techiques developed i previous chapters ca be applied to three differet kids of data. 1. Cross-sectioal data. 2. Time series data.

More information

Chapter 6 Sampling Distributions

Chapter 6 Sampling Distributions Chapter 6 Samplig Distributios 1 I most experimets, we have more tha oe measuremet for ay give variable, each measuremet beig associated with oe radomly selected a member of a populatio. Hece we eed to

More information

Poincaré Problem for Nonlinear Elliptic Equations of Second Order in Unbounded Domains

Poincaré Problem for Nonlinear Elliptic Equations of Second Order in Unbounded Domains Advaces i Pure Mathematics 23 3 72-77 http://dxdoiorg/4236/apm233a24 Published Olie Jauary 23 (http://wwwscirporg/oural/apm) Poicaré Problem for Noliear Elliptic Equatios of Secod Order i Ubouded Domais

More information

Solutions for the Exam 9 January 2012

Solutions for the Exam 9 January 2012 Mastermath ad LNMB Course: Discrete Optimizatio Solutios for the Exam 9 Jauary 2012 Utrecht Uiversity, Educatorium, 15:15 18:15 The examiatio lasts 3 hours. Gradig will be doe before Jauary 23, 2012. Studets

More information

Chapter 6 Infinite Series

Chapter 6 Infinite Series Chapter 6 Ifiite Series I the previous chapter we cosidered itegrals which were improper i the sese that the iterval of itegratio was ubouded. I this chapter we are goig to discuss a topic which is somewhat

More information

Linear Programming and the Simplex Method

Linear Programming and the Simplex Method Liear Programmig ad the Simplex ethod Abstract This article is a itroductio to Liear Programmig ad usig Simplex method for solvig LP problems i primal form. What is Liear Programmig? Liear Programmig is

More information

ROLL CUTTING PROBLEMS UNDER STOCHASTIC DEMAND

ROLL CUTTING PROBLEMS UNDER STOCHASTIC DEMAND Pacific-Asia Joural of Mathematics, Volume 5, No., Jauary-Jue 20 ROLL CUTTING PROBLEMS UNDER STOCHASTIC DEMAND SHAKEEL JAVAID, Z. H. BAKHSHI & M. M. KHALID ABSTRACT: I this paper, the roll cuttig problem

More information

Design and Analysis of Algorithms

Design and Analysis of Algorithms Desig ad Aalysis of Algorithms Probabilistic aalysis ad Radomized algorithms Referece: CLRS Chapter 5 Topics: Hirig problem Idicatio radom variables Radomized algorithms Huo Hogwei 1 The hirig problem

More information

Introduction to Optimization Techniques. How to Solve Equations

Introduction to Optimization Techniques. How to Solve Equations Itroductio to Optimizatio Techiques How to Solve Equatios Iterative Methods of Optimizatio Iterative methods of optimizatio Solutio of the oliear equatios resultig form a optimizatio problem is usually

More information

Regression with quadratic loss

Regression with quadratic loss Regressio with quadratic loss Maxim Ragisky October 13, 2015 Regressio with quadratic loss is aother basic problem studied i statistical learig theory. We have a radom couple Z = X,Y, where, as before,

More information

On forward improvement iteration for stopping problems

On forward improvement iteration for stopping problems O forward improvemet iteratio for stoppig problems Mathematical Istitute, Uiversity of Kiel, Ludewig-Mey-Str. 4, D-24098 Kiel, Germay irle@math.ui-iel.de Albrecht Irle Abstract. We cosider the optimal

More information

Introduction to Machine Learning DIS10

Introduction to Machine Learning DIS10 CS 189 Fall 017 Itroductio to Machie Learig DIS10 1 Fu with Lagrage Multipliers (a) Miimize the fuctio such that f (x,y) = x + y x + y = 3. Solutio: The Lagragia is: L(x,y,λ) = x + y + λ(x + y 3) Takig

More information

Vector Quantization: a Limiting Case of EM

Vector Quantization: a Limiting Case of EM . Itroductio & defiitios Assume that you are give a data set X = { x j }, j { 2,,, }, of d -dimesioal vectors. The vector quatizatio (VQ) problem requires that we fid a set of prototype vectors Z = { z

More information

Daniel Lee Muhammad Naeem Chingyu Hsu

Daniel Lee Muhammad Naeem Chingyu Hsu omplexity Aalysis of Optimal Statioary all Admissio Policy ad Fixed Set Partitioig Policy for OVSF-DMA ellular Systems Daiel Lee Muhammad Naeem higyu Hsu Backgroud Presetatio Outlie System Model all Admissio

More information

FLOWSHOP SCHEDULING USING A NETWORK APPROACH

FLOWSHOP SCHEDULING USING A NETWORK APPROACH ,*, A. M. H. Oladeide 1,*, A. I. Momodu 2 ad C. A. Oladeide 3 1, 2, 3DEPARTMENT OF PRODUCTION ENGINEERING, FACULTY OF ENGINEERING, UNIVERSITY OF BENIN, NIGERIA. E-mail addresses: 1 moladeide@uibe.edu,

More information

Lecture 11: Pseudorandom functions

Lecture 11: Pseudorandom functions COM S 6830 Cryptography Oct 1, 2009 Istructor: Rafael Pass 1 Recap Lecture 11: Pseudoradom fuctios Scribe: Stefao Ermo Defiitio 1 (Ge, Ec, Dec) is a sigle message secure ecryptio scheme if for all uppt

More information

On Random Line Segments in the Unit Square

On Random Line Segments in the Unit Square O Radom Lie Segmets i the Uit Square Thomas A. Courtade Departmet of Electrical Egieerig Uiversity of Califoria Los Ageles, Califoria 90095 Email: tacourta@ee.ucla.edu I. INTRODUCTION Let Q = [0, 1] [0,

More information

Application to Random Graphs

Application to Random Graphs A Applicatio to Radom Graphs Brachig processes have a umber of iterestig ad importat applicatios. We shall cosider oe of the most famous of them, the Erdős-Réyi radom graph theory. 1 Defiitio A.1. Let

More information

(A sequence also can be thought of as the list of function values attained for a function f :ℵ X, where f (n) = x n for n 1.) x 1 x N +k x N +4 x 3

(A sequence also can be thought of as the list of function values attained for a function f :ℵ X, where f (n) = x n for n 1.) x 1 x N +k x N +4 x 3 MATH 337 Sequeces Dr. Neal, WKU Let X be a metric space with distace fuctio d. We shall defie the geeral cocept of sequece ad limit i a metric space, the apply the results i particular to some special

More information

IJSER 1 INTRODUCTION. limitations with a large number of jobs and when the number of machines are more than two.

IJSER 1 INTRODUCTION. limitations with a large number of jobs and when the number of machines are more than two. Iteratioal Joural of Scietific & Egieerig Research, Volume 7, Issue, March-206 5 ISSN 2229-558 Schedulig to Miimize Maespa o Idetical Parallel Machies Yousef Germa*, Ibrahim Badi*, Ahmed Bair*, Ali Shetwa**

More information

Empirical Process Theory and Oracle Inequalities

Empirical Process Theory and Oracle Inequalities Stat 928: Statistical Learig Theory Lecture: 10 Empirical Process Theory ad Oracle Iequalities Istructor: Sham Kakade 1 Risk vs Risk See Lecture 0 for a discussio o termiology. 2 The Uio Boud / Boferoi

More information

Feedback in Iterative Algorithms

Feedback in Iterative Algorithms Feedback i Iterative Algorithms Charles Byre (Charles Byre@uml.edu), Departmet of Mathematical Scieces, Uiversity of Massachusetts Lowell, Lowell, MA 01854 October 17, 2005 Abstract Whe the oegative system

More information

Infinite Sequences and Series

Infinite Sequences and Series Chapter 6 Ifiite Sequeces ad Series 6.1 Ifiite Sequeces 6.1.1 Elemetary Cocepts Simply speakig, a sequece is a ordered list of umbers writte: {a 1, a 2, a 3,...a, a +1,...} where the elemets a i represet

More information

ON POINTWISE BINOMIAL APPROXIMATION

ON POINTWISE BINOMIAL APPROXIMATION Iteratioal Joural of Pure ad Applied Mathematics Volume 71 No. 1 2011, 57-66 ON POINTWISE BINOMIAL APPROXIMATION BY w-functions K. Teerapabolar 1, P. Wogkasem 2 Departmet of Mathematics Faculty of Sciece

More information

4. Partial Sums and the Central Limit Theorem

4. Partial Sums and the Central Limit Theorem 1 of 10 7/16/2009 6:05 AM Virtual Laboratories > 6. Radom Samples > 1 2 3 4 5 6 7 4. Partial Sums ad the Cetral Limit Theorem The cetral limit theorem ad the law of large umbers are the two fudametal theorems

More information

Research Article A New Second-Order Iteration Method for Solving Nonlinear Equations

Research Article A New Second-Order Iteration Method for Solving Nonlinear Equations Abstract ad Applied Aalysis Volume 2013, Article ID 487062, 4 pages http://dx.doi.org/10.1155/2013/487062 Research Article A New Secod-Order Iteratio Method for Solvig Noliear Equatios Shi Mi Kag, 1 Arif

More information

Riesz-Fischer Sequences and Lower Frame Bounds

Riesz-Fischer Sequences and Lower Frame Bounds Zeitschrift für Aalysis ud ihre Aweduge Joural for Aalysis ad its Applicatios Volume 1 (00), No., 305 314 Riesz-Fischer Sequeces ad Lower Frame Bouds P. Casazza, O. Christese, S. Li ad A. Lider Abstract.

More information

Ellipsoid Method for Linear Programming made simple

Ellipsoid Method for Linear Programming made simple Ellipsoid Method for Liear Programmig made simple Sajeev Saxea Dept. of Computer Sciece ad Egieerig, Idia Istitute of Techology, Kapur, INDIA-08 06 December 3, 07 Abstract I this paper, ellipsoid method

More information

Scheduling under Uncertainty using MILP Sensitivity Analysis

Scheduling under Uncertainty using MILP Sensitivity Analysis Schedulig uder Ucertaity usig MILP Sesitivity Aalysis M. Ierapetritou ad Zheya Jia Departmet of Chemical & Biochemical Egieerig Rutgers, the State Uiversity of New Jersey Piscataway, NJ Abstract The aim

More information

Chapter 7 Isoperimetric problem

Chapter 7 Isoperimetric problem Chapter 7 Isoperimetric problem Recall that the isoperimetric problem (see the itroductio its coectio with ido s proble) is oe of the most classical problem of a shape optimizatio. It ca be formulated

More information

Integer Programming (IP)

Integer Programming (IP) Iteger Programmig (IP) The geeral liear mathematical programmig problem where Mied IP Problem - MIP ma c T + h Z T y A + G y + y b R p + vector of positive iteger variables y vector of positive real variables

More information

Sequences and Series of Functions

Sequences and Series of Functions Chapter 6 Sequeces ad Series of Fuctios 6.1. Covergece of a Sequece of Fuctios Poitwise Covergece. Defiitio 6.1. Let, for each N, fuctio f : A R be defied. If, for each x A, the sequece (f (x)) coverges

More information

Simulation of Discrete Event Systems

Simulation of Discrete Event Systems Simulatio of Discrete Evet Systems Uit 9 Queueig Models Fall Witer 2014/2015 Uiv.-Prof. Dr.-Ig. Dipl.-Wirt.-Ig. Christopher M. Schlick Chair ad Istitute of Idustrial Egieerig ad Ergoomics RWTH Aache Uiversity

More information

SRC Technical Note June 17, Tight Thresholds for The Pure Literal Rule. Michael Mitzenmacher. d i g i t a l

SRC Technical Note June 17, Tight Thresholds for The Pure Literal Rule. Michael Mitzenmacher. d i g i t a l SRC Techical Note 1997-011 Jue 17, 1997 Tight Thresholds for The Pure Literal Rule Michael Mitzemacher d i g i t a l Systems Research Ceter 130 Lytto Aveue Palo Alto, Califoria 94301 http://www.research.digital.com/src/

More information

NICK DUFRESNE. 1 1 p(x). To determine some formulas for the generating function of the Schröder numbers, r(x) = a(x) =

NICK DUFRESNE. 1 1 p(x). To determine some formulas for the generating function of the Schröder numbers, r(x) = a(x) = AN INTRODUCTION TO SCHRÖDER AND UNKNOWN NUMBERS NICK DUFRESNE Abstract. I this article we will itroduce two types of lattice paths, Schröder paths ad Ukow paths. We will examie differet properties of each,

More information

NYU Center for Data Science: DS-GA 1003 Machine Learning and Computational Statistics (Spring 2018)

NYU Center for Data Science: DS-GA 1003 Machine Learning and Computational Statistics (Spring 2018) NYU Ceter for Data Sciece: DS-GA 003 Machie Learig ad Computatioal Statistics (Sprig 208) Brett Berstei, David Roseberg, Be Jakubowski Jauary 20, 208 Istructios: Followig most lab ad lecture sectios, we

More information

EE 4TM4: Digital Communications II Probability Theory

EE 4TM4: Digital Communications II Probability Theory 1 EE 4TM4: Digital Commuicatios II Probability Theory I. RANDOM VARIABLES A radom variable is a real-valued fuctio defied o the sample space. Example: Suppose that our experimet cosists of tossig two fair

More information

ECE 901 Lecture 12: Complexity Regularization and the Squared Loss

ECE 901 Lecture 12: Complexity Regularization and the Squared Loss ECE 90 Lecture : Complexity Regularizatio ad the Squared Loss R. Nowak 5/7/009 I the previous lectures we made use of the Cheroff/Hoeffdig bouds for our aalysis of classifier errors. Hoeffdig s iequality

More information

1 of 7 7/16/2009 6:06 AM Virtual Laboratories > 6. Radom Samples > 1 2 3 4 5 6 7 6. Order Statistics Defiitios Suppose agai that we have a basic radom experimet, ad that X is a real-valued radom variable

More information

Lecture 14: Graph Entropy

Lecture 14: Graph Entropy 15-859: Iformatio Theory ad Applicatios i TCS Sprig 2013 Lecture 14: Graph Etropy March 19, 2013 Lecturer: Mahdi Cheraghchi Scribe: Euiwoog Lee 1 Recap Bergma s boud o the permaet Shearer s Lemma Number

More information

Lecture 15: Learning Theory: Concentration Inequalities

Lecture 15: Learning Theory: Concentration Inequalities STAT 425: Itroductio to Noparametric Statistics Witer 208 Lecture 5: Learig Theory: Cocetratio Iequalities Istructor: Ye-Chi Che 5. Itroductio Recall that i the lecture o classificatio, we have see that

More information

Lecture 11 and 12: Basic estimation theory

Lecture 11 and 12: Basic estimation theory Lecture ad 2: Basic estimatio theory Sprig 202 - EE 94 Networked estimatio ad cotrol Prof. Kha March 2 202 I. MAXIMUM-LIKELIHOOD ESTIMATORS The maximum likelihood priciple is deceptively simple. Louis

More information

Classification of problem & problem solving strategies. classification of time complexities (linear, logarithmic etc)

Classification of problem & problem solving strategies. classification of time complexities (linear, logarithmic etc) Classificatio of problem & problem solvig strategies classificatio of time complexities (liear, arithmic etc) Problem subdivisio Divide ad Coquer strategy. Asymptotic otatios, lower boud ad upper boud:

More information

Supplemental Material: Proofs

Supplemental Material: Proofs Proof to Theorem Supplemetal Material: Proofs Proof. Let be the miimal umber of traiig items to esure a uique solutio θ. First cosider the case. It happes if ad oly if θ ad Rak(A) d, which is a special

More information

The multi capacitated clustering problem

The multi capacitated clustering problem The multi capacitated clusterig problem Bruo de Aayde Prata 1 Federal Uiversity of Ceará, Brazil Abstract Clusterig problems are combiatorial optimizatio problems wi several idustrial applicatios. The

More information

if j is a neighbor of i,

if j is a neighbor of i, We see that if i = j the the coditio is trivially satisfied. Otherwise, T ij (i) = (i)q ij mi 1, (j)q ji, ad, (i)q ij T ji (j) = (j)q ji mi 1, (i)q ij. (j)q ji Now there are two cases, if (j)q ji (i)q

More information

Lecture #20. n ( x p i )1/p = max

Lecture #20. n ( x p i )1/p = max COMPSCI 632: Approximatio Algorithms November 8, 2017 Lecturer: Debmalya Paigrahi Lecture #20 Scribe: Yua Deg 1 Overview Today, we cotiue to discuss about metric embeddigs techique. Specifically, we apply

More information

Linear Regression Models

Linear Regression Models Liear Regressio Models Dr. Joh Mellor-Crummey Departmet of Computer Sciece Rice Uiversity johmc@cs.rice.edu COMP 528 Lecture 9 15 February 2005 Goals for Today Uderstad how to Use scatter diagrams to ispect

More information

Support vector machine revisited

Support vector machine revisited 6.867 Machie learig, lecture 8 (Jaakkola) 1 Lecture topics: Support vector machie ad kerels Kerel optimizatio, selectio Support vector machie revisited Our task here is to first tur the support vector

More information

Chapter 2 The Monte Carlo Method

Chapter 2 The Monte Carlo Method Chapter 2 The Mote Carlo Method The Mote Carlo Method stads for a broad class of computatioal algorithms that rely o radom sampligs. It is ofte used i physical ad mathematical problems ad is most useful

More information

Economics 241B Relation to Method of Moments and Maximum Likelihood OLSE as a Maximum Likelihood Estimator

Economics 241B Relation to Method of Moments and Maximum Likelihood OLSE as a Maximum Likelihood Estimator Ecoomics 24B Relatio to Method of Momets ad Maximum Likelihood OLSE as a Maximum Likelihood Estimator Uder Assumptio 5 we have speci ed the distributio of the error, so we ca estimate the model parameters

More information

Chapter 4. Fourier Series

Chapter 4. Fourier Series Chapter 4. Fourier Series At this poit we are ready to ow cosider the caoical equatios. Cosider, for eample the heat equatio u t = u, < (4.) subject to u(, ) = si, u(, t) = u(, t) =. (4.) Here,

More information

Posted-Price, Sealed-Bid Auctions

Posted-Price, Sealed-Bid Auctions Posted-Price, Sealed-Bid Auctios Professors Greewald ad Oyakawa 207-02-08 We itroduce the posted-price, sealed-bid auctio. This auctio format itroduces the idea of approximatios. We describe how well this

More information

subject to A 1 x + A 2 y b x j 0, j = 1,,n 1 y j = 0 or 1, j = 1,,n 2

subject to A 1 x + A 2 y b x j 0, j = 1,,n 1 y j = 0 or 1, j = 1,,n 2 Additioal Brach ad Boud Algorithms 0-1 Mixed-Iteger Liear Programmig The brach ad boud algorithm described i the previous sectios ca be used to solve virtually all optimizatio problems cotaiig iteger variables,

More information

Polynomial Functions and Their Graphs

Polynomial Functions and Their Graphs Polyomial Fuctios ad Their Graphs I this sectio we begi the study of fuctios defied by polyomial expressios. Polyomial ad ratioal fuctios are the most commo fuctios used to model data, ad are used extesively

More information

Spectral Partitioning in the Planted Partition Model

Spectral Partitioning in the Planted Partition Model Spectral Graph Theory Lecture 21 Spectral Partitioig i the Plated Partitio Model Daiel A. Spielma November 11, 2009 21.1 Itroductio I this lecture, we will perform a crude aalysis of the performace of

More information

Information Theory and Statistics Lecture 4: Lempel-Ziv code

Information Theory and Statistics Lecture 4: Lempel-Ziv code Iformatio Theory ad Statistics Lecture 4: Lempel-Ziv code Łukasz Dębowski ldebowsk@ipipa.waw.pl Ph. D. Programme 203/204 Etropy rate is the limitig compressio rate Theorem For a statioary process (X i)

More information

Relations Among Algebras

Relations Among Algebras Itroductio to leee Algebra Lecture 6 CS786 Sprig 2004 February 9, 2004 Relatios Amog Algebras The otio of free algebra described i the previous lecture is a example of a more geeral pheomeo called adjuctio.

More information

Element sampling: Part 2

Element sampling: Part 2 Chapter 4 Elemet samplig: Part 2 4.1 Itroductio We ow cosider uequal probability samplig desigs which is very popular i practice. I the uequal probability samplig, we ca improve the efficiecy of the resultig

More information

DETERMINATION OF MECHANICAL PROPERTIES OF A NON- UNIFORM BEAM USING THE MEASUREMENT OF THE EXCITED LONGITUDINAL ELASTIC VIBRATIONS.

DETERMINATION OF MECHANICAL PROPERTIES OF A NON- UNIFORM BEAM USING THE MEASUREMENT OF THE EXCITED LONGITUDINAL ELASTIC VIBRATIONS. ICSV4 Cairs Australia 9- July 7 DTRMINATION OF MCHANICAL PROPRTIS OF A NON- UNIFORM BAM USING TH MASURMNT OF TH XCITD LONGITUDINAL LASTIC VIBRATIONS Pavel Aokhi ad Vladimir Gordo Departmet of the mathematics

More information

Dynamic Policy Programming with Function Approximation: Supplementary Material

Dynamic Policy Programming with Function Approximation: Supplementary Material Dyamic Policy Programmig with Fuctio pproximatio: Supplemetary Material Mohammad Gheshlaghi zar Radboud Uiversity Nijmege Geert Grooteplei Noord 21 6525 EZ Nijmege Netherlads m.azar@sciece.ru.l Viceç Gómez

More information

b i u x i U a i j u x i u x j

b i u x i U a i j u x i u x j M ath 5 2 7 Fall 2 0 0 9 L ecture 1 9 N ov. 1 6, 2 0 0 9 ) S ecod- Order Elliptic Equatios: Weak S olutios 1. Defiitios. I this ad the followig two lectures we will study the boudary value problem Here

More information

Machine Learning for Data Science (CS 4786)

Machine Learning for Data Science (CS 4786) Machie Learig for Data Sciece CS 4786) Lecture & 3: Pricipal Compoet Aalysis The text i black outlies high level ideas. The text i blue provides simple mathematical details to derive or get to the algorithm

More information

11. FINITE FIELDS. Example 1: The following tables define addition and multiplication for a field of order 4.

11. FINITE FIELDS. Example 1: The following tables define addition and multiplication for a field of order 4. 11. FINITE FIELDS 11.1. A Field With 4 Elemets Probably the oly fiite fields which you ll kow about at this stage are the fields of itegers modulo a prime p, deoted by Z p. But there are others. Now although

More information