Flows. Chapter Circulations
|
|
- Ira Daniel
- 5 years ago
- Views:
Transcription
1 Chapter 4 Flows For a directed graph D = (V,A), we define δ + (U) := {(u,v) A : u U,v / U} as the arcs leaving U and δ (U) := {(u,v) A u / U,v U} as the arcs entering U. 4. Circulations In a directed graph D = (V,A), we call a function f : A R 0 a circulation if f(δ (v)) = f(δ + (v)) v V. In some sense, circulations are flows that have no source and no sink. Let l : A R 0 be lower bounds and let u : A R 0 be upper bounds, then we say that a circulation f respects these bounds if l(a) f(a) u(a) for all a A. Given such bounds, we can again define a residual graph D f := (V,A f ) with A f := {a A f(a) < u(a)} {a a A : l(a) < f(a)}. Given such lower and upper bounds, it is not obvious which conditions should be satisfied so that a feasible circulation even exists. One obvious condition is that for every cut U, the lower bound on the incoming flow should not exceed the upper bound on the outgoing flow. It turns out that this condition is also sufficient: Theorem 7 (Hoffman s Circulation Theorem). Let D = (V,A) be a directed graph with l(a) u(a) a A. Then there exists a circulation f with l(a) f(a) u(a) for all a A if and only if l(δ (U)) u(δ + (U)) U V. Proof.. For any cut U and any a circulation f with l(a) f(a) u(a) one has l(δ (U)) l(a) f(a) f(δ (U)) circulation = f(δ + (U)) f(a) u(a) u(δ + (U)).. For any function f (not necessarily a circulation), define loss f (v) := f(δ + (v)) f(δ (v)). By compactness, there is a function f A R 0 with l(a) f(a) u(a) for a A minimizing v V loss f(v). Fix this choice of f. Define S := {v V : loss f (v) < 0} and T := {v V : loss f (v) > 0}. Suppose that S, since otherwise S = T = and f is a circulation. If there is a path P in the residual graph D f from a node in S to a node in T, then f can be augmented along that 5
2 path decreasing loss f. Hence we can assume that there is no S-T path. Define U := {v V : v reachable from S in D f }. U f(a) = u(a) S T f(a) = l(a) Then A f has no arcs leaving U. Hence for a δ + (U) one has f(a) = u(a) and a δ (U) one has f(a) = l(a). Then u(δ + (U)) l(δ (U)) = f(δ + (U)) f(δ (U)) loss f(v)=0 v U\S = loss f (S) < 0 That means U is a cut violating the condition. 4. Min Cost Circulations In this section we will discuss an algorithm solving the following problem: Min Cost Circulation: Given a directed graph D = (V,A) with edge cost c : A R, lower bounds l : A R 0 and upper bounds u : A R 0, find a circulation f : A R with l(a) f(a) u(a) for all a A minimizing c(f) := a A c(a) f(a). It is not difficult to model for example the maximum s-t flow problem or the minimum cost max s-t flow problem as a min cost circulation. 4.. The residual graph Now that our circulation also includes cost, we need to extend the definition of a residual graph. For a directed D = (V,A) with edge cost c : A R, lower bounds l : A R 0 and upper bounds u : A R 0, we define the residual graph D f = (V,A f ) that contains the following two types of arcs: For a A with f(a) < u(a) we have a A f with residual capacity u f (a) := u(a) f(a) and residual cost c(a). For a A with f(a) > l(a) we have a A f with residual capacity u f (a ) := f(a) l(a) and residual cost c(a ) := c(a). Observation 8. Let f : A R 0 be a circulation with l(a) f(a) u(a) and let C A f be a directed circuit in the residual graph. Set := min{u f (a) : a C}. Then f : A R with f(a)+ if a C f (a) := f(a) if a C f(a) otherwise is a circulation with l(a) f (a) u(a) and c(f ) = c(f)+ c(c). 6
3 4.. Flow decomposition Let us call a circulation atomic if there is a single directed circuit C A so that { if a C f(a) = 0 otherwise for some 0. The following technique is also called flow decomposition in the literature: Lemma 9. Let f : A R 0 be a circulation in D = (V,A). Then there are atomic circulations f,...,f k : A R 0 with f = f +...+f k and k A. Proof. Every node with positive in-degree in {a A f(a) > 0} has also positive out-degree. Find any simple cycle C {a A f(a) > 0}. Choose := min{f(a) : a C} and define an atomic circulation f with circuit C and value. Repeat with the non-negative circulation f f. In each iteration the flow of at least one arc goes down to 0, which implies that at most A iterations are needed Optimality criterion It turns out that optimality of a circulation is equivalent to not having a negative cost cycle in the residual graph: Lemma 0. Let f be a circulation in D = (V,A) with l(a) f(a) u(a) for all a A. Then f is not optimal there is a negative cost cycle in D f. Proof. If D f contains a negative cost cycle C, then augmenting the flow on f decreases the cost, see Observation 8.. Let f be an optimum circulation. Define f : A f R 0 by { f f (a) f(a) if f(a) < f (a) u(a) (a) := f(a ) f (a ) if f(a ) > f (a ) l(a ) Then f is a circulation with c(f ) = c(f ) c(f). f in solid; f in dashed f in solid Using Lemma 9 we can split f = k i= f i where f,...,f k : A f R 0 are atomic circulations. Then k i= c(f i) = c(f ) c(f) < 0, hence for some i we have c(f i ) < 0. The cycle corresponding to f i then has negative cost. 7
4 4..4 The algorithm The obvious idea is to augment the current circulation along a cycle with negative cost in D f. Already in case of maximum flows, we learned that the choice of a cycle should be done carefully in order to obtain a polynomial running time. It will turn out that the right cycle is the one that minimizes the average cost. Let us define µ(f) := { c(c) } min C cycle in D f C as the cost of the minimum mean cycle. Note that we allow the empty cycle so that always µ(f) 0. Lemma. Given a directed graph D = (V,A) with edge cost c : A R. Then a minimum mean length cycle can be found in time O( V A ). Proof. Consider the following slight variation of the Bellman-Ford algorithm: set d 0 (u,v) := { 0 if u = v if u v and d k (u,v) := min (w,v) A {d k (u,w)+c(w,v)} for k =,..., V and u,v V. Then d k (u,v) = c(p) where P is the u-v walk with exactly k arcs that minimizes c(p). Computing all entries takes time O( V A ). Then the cost of the minimum mean cycle is min k=0,..., V ; u V { dk (u,u) } (counting the entries for k = 0 as having value 0) and also the cycle itself can be recovered from the entries. Note that the minimum mean cost cycle can be found with a more sophisticated argument in time O( V A ), see for example Chapter 7.3 in the textbook of Korte and Vygen. The algorithm that will use for min cost circulation is the following: Minimum mean cycle canceling algorithm () Compute any feasible circulation f with l(a) f(a) u(a) for all a A. () WHILE cycle in D f with negative cost DO (3) Compute minimum mean cycle C A f (4) Augment f along C by min{u f (a) : a C} Step () can be implemented, by solving a min cost circulation problem in the graph D = (V,A ) which for each a = (u,v) A contains two parallel arcs a,a from u to v with parameters: c(a ) =, l(a ) = 0, u(a ) = l(a) c(a ) = 0, l(a ) = 0, u(a ) = u(a) l(a) Then f = 0 is a feasible circulation in D and any mincost circulation in D corresponds exactly to a feasible circulation in D. 8 k
5 4..5 Node potentials For a directed graph D = (V,A) with edge cost c : A R, a function π : V R are called node potentials. These induce reduced costs c π i,j := c ij +π i π j. Lemma. Let D = (V,A) be a directed graph with cost c : A R and node potentials π. Then any cycle C A has c π (C) = c(c). Proof. Clear as the node potentials cancel out. Lemma 3. Let D = (V,A) be a directed graph with arc cost c : A R. Then D has no negative cost cycle there are node potentials with c π i,j := c ij +π i π j 0 (i,j) A. Proof.. Any cycle C has c(c) = c π (C) 0 by Lemma. We add an artificial node s and arcs (s, v) with c(s, v) := 0 for all v V. s D Let d(u,v) be the length of the shortest path u-v in the resulting graph. Note that d(u,v) maybe negative, but since there are no negative cost cycle, d is well defined. Define π(i) := d(s,i) as the node potentials. Then c π i,j = c ij +d(s,i) d(s,j) 0 d(s,j) d(s,i)+c ij which is the triangle inequality. A circulation f : A R 0 is ε-optimal for ε 0 if there are node potentials π : V R so that c π ij for all (i,j) A f. Let ε(f) := min{ε 0 : f is ε-optimal} The interpretation of this quantity is that ε(f) is the smallest amount that has to be added to the cost of the arcs in the residual graph to eliminate all negative cost cycles. Lemma 4. For any circulation f, µ(f) = (f). Proof. Let π be the node potentials valid for ε(f). The minimum mean cycle C has C µ(f) = c(c) = c π (C) (f) C.. Let use define a cost function c(u,v) := c(u,v) µ(f). Now there is no negative cost cycle w.r.t. c and there are node potentials π with c(i,j)+π i π j 0, which is the same as c(i,j)+π i π j µ(f) The main analysis The analysis of the cycle cancelling method will consist of showing that the value ofε(f) is decreasing in the course of the algorithm. Lemma 5. Update f to f by augmenting along a minimum mean cost cycle. Then ε(f ) ε(f). 9
6 Proof. Abbreviate ε := ε(f). Let π : V R be the node potentials with c π (a) for every arc a A f. We will show that the same node potentials are still feasible for the updated graph D f. Let C A f be the minimum cost mean cycle. Note that its cost are c π (C) = C ε by Lemma 4. That means that every arc a C must have c π (a) =. The only new arcs (i,j) A f \A f have (j,i) C. Hence the reduced cost are c π ij = c ji +π i π j = c π ji = ε 0. We should remark that the claim of Lemma 5 is false if an arbitrary cycle is chosen for updates. To understand the final part of the analysis better, fix a current circulation f with ε := µ(f) and let π be the corresponding potential. Suppose for the sake of argument that the next two mean cycles are C and C contains the reverse of the arc that was the bottleneck for C. Then the picture for the reduced cost in the residual graph will be as follows: +ε cycle C cycle C But then the mean cost of C has to be higher as it contains an arc with reduced cost +ε. This idea can be generalized as follows: Lemma 6. Consider a sequence {f i } i 0 of circulations where f i+ emerges from f i by augmenting along the minimum mean cycle in D fi. Then ε(f A + ) ( V ) ε(f 0). Proof. Let ε := ε(f 0 ) and fix the potentials π : V R with c π (a) for all a A f0. Claim I. There is a k {0,..., A + } so that the minimum mean cost cycle C D fk in that iteration contains an arc a C with c π (a) 0. Proof of claim. As long as the current iteration k uses a minimum mean cost cycle C D fk with c π (a) < 0, every arc that appears new in the residual graph will non-negative reduced cost. Moreover, in every iteration at least one arc is bottleneck arc and will not appear in the residual graph of the next iteration. This can only go on for at most A iterations. Claim II. Consider the minimal such k from Claim I. Then µ(f k ) ( V ) ε. Proof of claim. Let C A fk be the minimum mean cycle. We know that c π (a) a A fk. Then c(c) = c π (C) ( C ) () and hence µ(f k ) = c(c) C ( C ) (). The claim follows from C V. Theorem 7. Suppose the cost function is c : A { c max,...,+c max }. Then the minimum mean cycle cancelling algorithm terminates after V ( A +) ln( V c max )) iterations. Proof. Let f be the flow before the st iteration and f is the circulation after V ( A +) ln( V c max ) iterations. Then f is c max -optimal. Moreover, ( ε(f ) c max ) V ln( V cmax) cmax exp( ln( V c max )) = V V. Since the cost are integral, that implies that actually ε(f ) = 0. 0
7 Surprisingly one can prove that the number of iterations is bounded by a polynomial only in V and A, indpendent from cost and capacity values. Theorem 8 (Tardos 985). The minimum mean cost cycle algorithm terminates aftero( A V ln( V )) iterations. We will not give a proof here. Exercises Exercise 4.. Let D = (V,A) be a directed graph with incidence matrix M D {,0,} V A. For a set of arcs C A, we write χ C {0,} A as the characteristic vector. Consider P = {x R A M D x = 0; x 0} and Q = cone{χ C : C A is a circuit}. Prove that P = Q. Exercise 4.. Use Hoffman s Circulation Theorem to prove the non-trivial direction of the MaxFlow=MinCut Theorem using a suitable reduction. Recall that the non-trivial direction says that for a directed graph D = (V,A) with s,t V, capacities u(a) 0 for a A and a maximum flow value of F, there is an s-t cut of value F. Exercise 4.3. Let D = (V,A) be a directed graph with lower and upper bounds l,u : A R 0 and cost c : A R. Let f,f be circulations with c(f ) < c(f). Show that there exists a cycle C A f in the residual graph with bottleneck capacity α := min{u f (a) : a C} so that α c(c) A (c(f ) c(f)).
Introduction to Algorithms
Introduction to Algorithms 6.046J/18.401J/SMA5503 Lecture 18 Prof. Erik Demaine Negative-weight cycles Recall: If a graph G = (V, E) contains a negativeweight cycle, then some shortest paths may not exist.
More informationdirected weighted graphs as flow networks the Ford-Fulkerson algorithm termination and running time
Network Flow 1 The Maximum-Flow Problem directed weighted graphs as flow networks the Ford-Fulkerson algorithm termination and running time 2 Maximum Flows and Minimum Cuts flows and cuts max flow equals
More informationThe min cost flow problem Course notes for Optimization Spring 2007
The min cost flow problem Course notes for Optimization Spring 2007 Peter Bro Miltersen February 7, 2007 Version 3.0 1 Definition of the min cost flow problem We shall consider a generalization of the
More informationRunning Time. Assumption. All capacities are integers between 1 and C.
Running Time Assumption. All capacities are integers between and. Invariant. Every flow value f(e) and every residual capacities c f (e) remains an integer throughout the algorithm. Theorem. The algorithm
More informationMath 5490 Network Flows
Math 90 Network Flows Lecture 8: Flow Decomposition Algorithm Stephen Billups University of Colorado at Denver Math 90Network Flows p./6 Flow Decomposition Algorithms Two approaches to modeling network
More informationMathematics for Decision Making: An Introduction. Lecture 13
Mathematics for Decision Making: An Introduction Lecture 13 Matthias Köppe UC Davis, Mathematics February 17, 2009 13 1 Reminder: Flows in networks General structure: Flows in networks In general, consider
More informationMaximum flow problem (part I)
Maximum flow problem (part I) Combinatorial Optimization Giovanni Righini Università degli Studi di Milano Definitions A flow network is a digraph D = (N,A) with two particular nodes s and t acting as
More informationLecture 2: Network Flows 1
Comp 260: Advanced Algorithms Tufts University, Spring 2011 Lecture by: Prof. Cowen Scribe: Saeed Majidi Lecture 2: Network Flows 1 A wide variety of problems, including the matching problems discussed
More informationWeek 4. (1) 0 f ij u ij.
Week 4 1 Network Flow Chapter 7 of the book is about optimisation problems on networks. Section 7.1 gives a quick introduction to the definitions of graph theory. In fact I hope these are already known
More informationThe maximum flow problem
The maximum flow problem A. Agnetis 1 Basic properties Given a network G = (N, A) (having N = n nodes and A = m arcs), and two nodes s (source) and t (sink), the maximum flow problem consists in finding
More informationWe say that a flow is feasible for G (or just feasible if the graph and capacity function in question are obvious) if
CS787: Advanced Algorithms Lecture 4: Network Flow We devote this lecture to the network flow problem and the max-flow min-cut theorem. A number of other problems can be cast into a maximum flow or minimum
More information1 Review for Lecture 2 MaxFlow
Comp 260: Advanced Algorithms Tufts University, Spring 2009 Prof. Lenore Cowen Scribe: Wanyu Wang Lecture 13: Back to MaxFlow/Edmonds-Karp 1 Review for Lecture 2 MaxFlow A flow network showing flow and
More information6.046 Recitation 11 Handout
6.046 Recitation 11 Handout May 2, 2008 1 Max Flow as a Linear Program As a reminder, a linear program is a problem that can be written as that of fulfilling an objective function and a set of constraints
More informationMaximum Integer Flows in Directed Planar Graphs with Multiple Sources and Sinks and Vertex Capacities
Maximum Integer Flows in Directed Planar Graphs with Multiple Sources and Sinks and Vertex Capacities Yipu Wang University of Illinois at Urbana-Champaign ywang298@illinois.edu July 12, 2018 Abstract We
More informationLecture 3. 1 Polynomial-time algorithms for the maximum flow problem
ORIE 633 Network Flows August 30, 2007 Lecturer: David P. Williamson Lecture 3 Scribe: Gema Plaza-Martínez 1 Polynomial-time algorithms for the maximum flow problem 1.1 Introduction Let s turn now to considering
More informationAlgorithms: COMP3121/3821/9101/9801
NEW SOUTH WALES Algorithms: COMP32/382/90/980 Aleks Ignjatović School of Computer Science and Engineering University of New South Wales LECTURE 7: MAXIMUM FLOW COMP32/382/90/980 / 24 Flow Networks A flow
More informationThe min cost flow problem Course notes for Search and Optimization Spring 2004
The min cost flow problem Course notes for Search and Optimization Spring 2004 Peter Bro Miltersen February 20, 2004 Version 1.3 1 Definition of the min cost flow problem We shall consider a generalization
More informationChapter 7 Network Flow Problems, I
Chapter 7 Network Flow Problems, I Network flow problems are the most frequently solved linear programming problems. They include as special cases, the assignment, transportation, maximum flow, and shortest
More informationCMPSCI 611: Advanced Algorithms
CMPSCI 611: Advanced Algorithms Lecture 12: Network Flow Part II Andrew McGregor Last Compiled: December 14, 2017 1/26 Definitions Input: Directed Graph G = (V, E) Capacities C(u, v) > 0 for (u, v) E and
More informationEnergy minimization via graph-cuts
Energy minimization via graph-cuts Nikos Komodakis Ecole des Ponts ParisTech, LIGM Traitement de l information et vision artificielle Binary energy minimization We will first consider binary MRFs: Graph
More informationTwo Applications of Maximum Flow
Two Applications of Maximum Flow The Bipartite Matching Problem a bipartite graph as a flow network maximum flow and maximum matching alternating paths perfect matchings 2 Circulation with Demands flows
More informationDiscrete Optimization 2010 Lecture 2 Matroids & Shortest Paths
Matroids Shortest Paths Discrete Optimization 2010 Lecture 2 Matroids & Shortest Paths Marc Uetz University of Twente m.uetz@utwente.nl Lecture 2: sheet 1 / 25 Marc Uetz Discrete Optimization Matroids
More informationFlows and Cuts. 1 Concepts. CS 787: Advanced Algorithms. Instructor: Dieter van Melkebeek
CS 787: Advanced Algorithms Flows and Cuts Instructor: Dieter van Melkebeek This lecture covers the construction of optimal flows and cuts in networks, their relationship, and some applications. It paves
More informationDynamic Programming: Shortest Paths and DFA to Reg Exps
CS 374: Algorithms & Models of Computation, Spring 207 Dynamic Programming: Shortest Paths and DFA to Reg Exps Lecture 8 March 28, 207 Chandra Chekuri (UIUC) CS374 Spring 207 / 56 Part I Shortest Paths
More informationDiscrete Optimization 2010 Lecture 3 Maximum Flows
Remainder: Shortest Paths Maximum Flows Discrete Optimization 2010 Lecture 3 Maximum Flows Marc Uetz University of Twente m.uetz@utwente.nl Lecture 3: sheet 1 / 29 Marc Uetz Discrete Optimization Outline
More informationFlow Network. The following figure shows an example of a flow network:
Maximum Flow 1 Flow Network The following figure shows an example of a flow network: 16 V 1 12 V 3 20 s 10 4 9 7 t 13 4 V 2 V 4 14 A flow network G = (V,E) is a directed graph. Each edge (u, v) E has a
More informationCMSC 451: Max-Flow Extensions
CMSC 51: Max-Flow Extensions Slides By: Carl Kingsford Department of Computer Science University of Maryland, College Park Based on Section 7.7 of Algorithm Design by Kleinberg & Tardos. Circulations with
More informationMulticommodity Flows and Column Generation
Lecture Notes Multicommodity Flows and Column Generation Marc Pfetsch Zuse Institute Berlin pfetsch@zib.de last change: 2/8/2006 Technische Universität Berlin Fakultät II, Institut für Mathematik WS 2006/07
More informationC&O 355 Mathematical Programming Fall 2010 Lecture 18. N. Harvey
C&O 355 Mathematical Programming Fall 2010 Lecture 18 N. Harvey Network Flow Topics Max Flow / Min Cut Theorem Total Unimodularity Directed Graphs & Incidence Matrices Proof of Max Flow / Min Cut Theorem
More informationMaximum flow problem CE 377K. February 26, 2015
Maximum flow problem CE 377K February 6, 05 REVIEW HW due in week Review Label setting vs. label correcting Bellman-Ford algorithm Review MAXIMUM FLOW PROBLEM Maximum Flow Problem What is the greatest
More informationDynamic Programming: Shortest Paths and DFA to Reg Exps
CS 374: Algorithms & Models of Computation, Fall 205 Dynamic Programming: Shortest Paths and DFA to Reg Exps Lecture 7 October 22, 205 Chandra & Manoj (UIUC) CS374 Fall 205 / 54 Part I Shortest Paths with
More information1 Integer Decomposition Property
CS 598CSC: Combinatorial Optimization Lecture date: Feb 2, 2010 Instructor: Chandra Chekuri Scribe: Siva Theja Maguluri Material taken mostly from [1] (Chapter 19). 1 Integer Decomposition Property A polyhedron
More informationComputer Science & Engineering 423/823 Design and Analysis of Algorithms
Bipartite Matching Computer Science & Engineering 423/823 Design and s Lecture 07 (Chapter 26) Stephen Scott (Adapted from Vinodchandran N. Variyam) 1 / 36 Spring 2010 Bipartite Matching 2 / 36 Can use
More informationMaximum Flow. Jie Wang. University of Massachusetts Lowell Department of Computer Science. J. Wang (UMass Lowell) Maximum Flow 1 / 27
Maximum Flow Jie Wang University of Massachusetts Lowell Department of Computer Science J. Wang (UMass Lowell) Maximum Flow 1 / 27 Flow Networks A flow network is a weighted digraph G = (V, E), where the
More informationGraphs and Network Flows IE411. Lecture 12. Dr. Ted Ralphs
Graphs and Network Flows IE411 Lecture 12 Dr. Ted Ralphs IE411 Lecture 12 1 References for Today s Lecture Required reading Sections 21.1 21.2 References AMO Chapter 6 CLRS Sections 26.1 26.2 IE411 Lecture
More information7. Lecture notes on the ellipsoid algorithm
Massachusetts Institute of Technology Michel X. Goemans 18.433: Combinatorial Optimization 7. Lecture notes on the ellipsoid algorithm The simplex algorithm was the first algorithm proposed for linear
More informationMinimum cost transportation problem
Minimum cost transportation problem Complements of Operations Research Giovanni Righini Università degli Studi di Milano Definitions The minimum cost transportation problem is a special case of the minimum
More informationAn 1.75 approximation algorithm for the leaf-to-leaf tree augmentation problem
An 1.75 approximation algorithm for the leaf-to-leaf tree augmentation problem Zeev Nutov, László A. Végh January 12, 2016 We study the tree augmentation problem: Tree Augmentation Problem (TAP) Instance:
More informationThe Max Flow Problem
The Max Flow Problem jla,jc@imm.dtu.dk Informatics and Mathematical Modelling Technical University of Denmark 1 1 3 4 6 6 1 2 4 3 r 3 2 4 5 7 s 6 2 1 8 1 3 3 2 6 2 Max-Flow Terminology We consider a digraph
More informationThe Steiner Network Problem
The Steiner Network Problem Pekka Orponen T-79.7001 Postgraduate Course on Theoretical Computer Science 7.4.2008 Outline 1. The Steiner Network Problem Linear programming formulation LP relaxation 2. The
More information2.13 Maximum flow with a strictly positive initial feasible flow
ex-.-. Foundations of Operations Research Prof. E. Amaldi. Maximum flow and minimum cut iven the following network with capacities on the arcs find a maximum (feasible) flow from node to node, and determine
More informationProblem set 1. (c) Is the Ford-Fulkerson algorithm guaranteed to produce an acyclic maximum flow?
CS261, Winter 2017. Instructor: Ashish Goel. Problem set 1 Electronic submission to Gradescope due 11:59pm Thursday 2/2. Form a group of 2-3 students that is, submit one homework with all of your names.
More informationACO Comprehensive Exam March 20 and 21, Computability, Complexity and Algorithms
1. Computability, Complexity and Algorithms Part a: You are given a graph G = (V,E) with edge weights w(e) > 0 for e E. You are also given a minimum cost spanning tree (MST) T. For one particular edge
More informationPartitioning Metric Spaces
Partitioning Metric Spaces Computational and Metric Geometry Instructor: Yury Makarychev 1 Multiway Cut Problem 1.1 Preliminaries Definition 1.1. We are given a graph G = (V, E) and a set of terminals
More informationCS 6820 Fall 2014 Lectures, October 3-20, 2014
Analysis of Algorithms Linear Programming Notes CS 6820 Fall 2014 Lectures, October 3-20, 2014 1 Linear programming The linear programming (LP) problem is the following optimization problem. We are given
More informationLecture 13: Polynomial-Time Algorithms for Min Cost Flows. (Reading: AM&O Chapter 10)
Lecture 1: Polynomial-Time Algorithms for Min Cost Flows (Reading: AM&O Chapter 1) Polynomial Algorithms for Min Cost Flows Improvements on the two algorithms for min cost flow: Successive Shortest Path
More informationORIE 633 Network Flows October 4, Lecture 10
ORIE 633 Network Flows October 4, 2007 Lecturer: David P. Williamson Lecture 10 Scribe: Kathleen King 1 Efficient algorithms for max flows 1.1 The Goldberg-Rao algorithm Recall from last time: Dinic s
More informationShortest paths with negative lengths
Chapter 8 Shortest paths with negative lengths In this chapter we give a linear-space, nearly linear-time algorithm that, given a directed planar graph G with real positive and negative lengths, but no
More informationMathematics for Decision Making: An Introduction. Lecture 8
Mathematics for Decision Making: An Introduction Lecture 8 Matthias Köppe UC Davis, Mathematics January 29, 2009 8 1 Shortest Paths and Feasible Potentials Feasible Potentials Suppose for all v V, there
More informationDiscrete Optimization
Prof. Friedrich Eisenbrand Martin Niemeier Due Date: April 15, 2010 Discussions: March 25, April 01 Discrete Optimization Spring 2010 s 3 You can hand in written solutions for up to two of the exercises
More information3.7 Cutting plane methods
3.7 Cutting plane methods Generic ILP problem min{ c t x : x X = {x Z n + : Ax b} } with m n matrix A and n 1 vector b of rationals. According to Meyer s theorem: There exists an ideal formulation: conv(x
More information5 Flows and cuts in digraphs
5 Flows and cuts in digraphs Recall that a digraph or network is a pair G = (V, E) where V is a set and E is a multiset of ordered pairs of elements of V, which we refer to as arcs. Note that two vertices
More information(1,3) (3,4) (2,2) (1,2) (2,4) t
Maximum flows Some catastrophe has happened in some far away place, and help is needed badly. Fortunately, all that is needed is available, but it is stored in some depot that is quite remote from the
More informationMaximum flow problem
Maximum flow problem 7000 Network flows Network Directed graph G = (V, E) Source node s V, sink node t V Edge capacities: cap : E R 0 Flow: f : E R 0 satisfying 1. Flow conservation constraints e:target(e)=v
More informationCSC 373: Algorithm Design and Analysis Lecture 12
CSC 373: Algorithm Design and Analysis Lecture 12 Allan Borodin February 4, 2013 1 / 16 Lecture 12: Announcements and Outline Announcements Term test 1 in tutorials. Need to use only two rooms due to sickness
More information15-850: Advanced Algorithms CMU, Fall 2018 HW #4 (out October 17, 2018) Due: October 28, 2018
15-850: Advanced Algorithms CMU, Fall 2018 HW #4 (out October 17, 2018) Due: October 28, 2018 Usual rules. :) Exercises 1. Lots of Flows. Suppose you wanted to find an approximate solution to the following
More informationLecture notes on the ellipsoid algorithm
Massachusetts Institute of Technology Handout 1 18.433: Combinatorial Optimization May 14th, 007 Michel X. Goemans Lecture notes on the ellipsoid algorithm The simplex algorithm was the first algorithm
More informationChapter 11. Approximation Algorithms. Slides by Kevin Wayne Pearson-Addison Wesley. All rights reserved.
Chapter 11 Approximation Algorithms Slides by Kevin Wayne. Copyright @ 2005 Pearson-Addison Wesley. All rights reserved. 1 Approximation Algorithms Q. Suppose I need to solve an NP-hard problem. What should
More information- Well-characterized problems, min-max relations, approximate certificates. - LP problems in the standard form, primal and dual linear programs
LP-Duality ( Approximation Algorithms by V. Vazirani, Chapter 12) - Well-characterized problems, min-max relations, approximate certificates - LP problems in the standard form, primal and dual linear programs
More information3.4 Relaxations and bounds
3.4 Relaxations and bounds Consider a generic Discrete Optimization problem z = min{c(x) : x X} with an optimal solution x X. In general, the algorithms generate not only a decreasing sequence of upper
More informationStrongly Polynomial Algorithm for a Class of Minimum-Cost Flow Problems with Separable Convex Objectives
Strongly Polynomial Algorithm for a Class of Minimum-Cost Flow Problems with Separable Convex Objectives László A. Végh April 12, 2013 Abstract A well-studied nonlinear extension of the minimum-cost flow
More informationLecture 2: Just married
COMP36111: Advanced Algorithms I Lecture 2: Just married Ian Pratt-Hartmann Room KB2.38: email: ipratt@cs.man.ac.uk 2017 18 Outline Matching Flow networks Third-year projects The stable marriage problem
More informationLecture 7: Shortest Paths in Graphs with Negative Arc Lengths. Reading: AM&O Chapter 5
Lecture 7: Shortest Paths in Graphs with Negative Arc Lengths Reading: AM&O Chapter Label Correcting Methods Assume the network G is allowed to have negative arc lengths but no directed negativelyweighted
More informationNetwork Flows made simple
Network Flows made simple A mild but rigorous introduction to Network Flows Periklis A. Papakonstantinou York University Network Flows is the last algorithmic paradigm we consider. Actually, we only address
More informationLecture 8 Network Optimization Algorithms
Advanced Algorithms Floriano Zini Free University of Bozen-Bolzano Faculty of Computer Science Academic Year 2013-2014 Lecture 8 Network Optimization Algorithms 1 21/01/14 Introduction Network models have
More informationRobust Network Codes for Unicast Connections: A Case Study
Robust Network Codes for Unicast Connections: A Case Study Salim Y. El Rouayheb, Alex Sprintson, and Costas Georghiades Department of Electrical and Computer Engineering Texas A&M University College Station,
More informationINVERSE SPANNING TREE PROBLEMS: FORMULATIONS AND ALGORITHMS
INVERSE SPANNING TREE PROBLEMS: FORMULATIONS AND ALGORITHMS P. T. Sokkalingam Department of Mathematics Indian Institute of Technology, Kanpur-208 016, INDIA Ravindra K. Ahuja Dept. of Industrial & Management
More informationExposition of the Dinitz algorithm
Exposition of the Dinitz algorithm Yefim Dinitz and Avraham A. Melkman Department of Computer Science Ben Gurion University of the Negev, 84105 Beer-Sheva, Israel {dinitz,melkman}@cs.bgu.ac.il June 6,
More information12. LOCAL SEARCH. gradient descent Metropolis algorithm Hopfield neural networks maximum cut Nash equilibria
12. LOCAL SEARCH gradient descent Metropolis algorithm Hopfield neural networks maximum cut Nash equilibria Lecture slides by Kevin Wayne Copyright 2005 Pearson-Addison Wesley h ttp://www.cs.princeton.edu/~wayne/kleinberg-tardos
More information6.854 Advanced Algorithms
6.854 Advanced Algorithms Homework 5 Solutions 1 10 pts Define the following sets: P = positions on the results page C = clicked old results U = unclicked old results N = new results and let π : (C U)
More informationChapter 11. Approximation Algorithms. Slides by Kevin Wayne Pearson-Addison Wesley. All rights reserved.
Chapter 11 Approximation Algorithms Slides by Kevin Wayne. Copyright @ 2005 Pearson-Addison Wesley. All rights reserved. 1 P and NP P: The family of problems that can be solved quickly in polynomial time.
More informationLecture 9 Tuesday, 4/20/10. Linear Programming
UMass Lowell Computer Science 91.503 Analysis of Algorithms Prof. Karen Daniels Spring, 2010 Lecture 9 Tuesday, 4/20/10 Linear Programming 1 Overview Motivation & Basics Standard & Slack Forms Formulating
More informationThe Budget-Constrained Maximum Flow Problem
9 The Budget-Constrained Maximum Flow Problem In this chapter we consider the following problem which is called the constrained maximum flow problem ( Cmfp) [AO95]: We are given a budget B and we seek
More informationApproximate Binary Search Algorithms for Mean Cuts and Cycles
Approximate Binary Search Algorithms for Mean Cuts and Cycles S. Thomas McCormick Faculty of Commerce and Business Administration University of British Columbia Vancouver, BC V6T 1Z2 Canada June 1992,
More informationAlgorithms and Theory of Computation. Lecture 11: Network Flow
Algorithms and Theory of Computation Lecture 11: Network Flow Xiaohui Bei MAS 714 September 18, 2018 Nanyang Technological University MAS 714 September 18, 2018 1 / 26 Flow Network A flow network is a
More informationIII. Linear Programming
III. Linear Programming Thomas Sauerwald Easter 2017 Outline Introduction Standard and Slack Forms Formulating Problems as Linear Programs Simplex Algorithm Finding an Initial Solution III. Linear Programming
More informationAlgorithm Design and Analysis
Algorithm Design and Analysis LETURE 2 Network Flow Finish bipartite matching apacity-scaling algorithm Adam Smith 0//0 A. Smith; based on slides by E. Demaine,. Leiserson, S. Raskhodnikova, K. Wayne Marriage
More informationInternet Routing Example
Internet Routing Example Acme Routing Company wants to route traffic over the internet from San Fransisco to New York. It owns some wires that go between San Francisco, Houston, Chicago and New York. The
More informationAlgorithm Design and Analysis
Algorithm Design and Analysis LETURE 28 Network Flow hoosing good augmenting paths apacity scaling algorithm A. Smith /4/2008 A. Smith; based on slides by K. Wayne and S. Raskhodnikova Pre-break: Ford-Fulkerson
More informationAlgorithms: Lecture 12. Chalmers University of Technology
Algorithms: Lecture 1 Chalmers University of Technology Today s Topics Shortest Paths Network Flow Algorithms Shortest Path in a Graph Shortest Path Problem Shortest path network. Directed graph G = (V,
More informationMax Flow: Algorithms and Applications
Max Flow: Algorithms and Applications Outline for today Quick review of the Max Flow problem Applications of Max Flow The Ford-Fulkerson (FF) algorithm for Max Flow Analysis of FF, and the Max Flow Min
More informationSolving the MWT. Recall the ILP for the MWT. We can obtain a solution to the MWT problem by solving the following ILP:
Solving the MWT Recall the ILP for the MWT. We can obtain a solution to the MWT problem by solving the following ILP: max subject to e i E ω i x i e i C E x i {0, 1} x i C E 1 for all critical mixed cycles
More informationAn Introduction to Network Flows Over Time
An Introduction to Network Flows Over Time Martin Skutella Abstract We give an introduction into the fascinating area of flows over time also called dynamic flows in the literature. Starting from the early
More informationCS 4407 Algorithms Lecture: Shortest Path Algorithms
CS 440 Algorithms Lecture: Shortest Path Algorithms Prof. Gregory Provan Department of Computer Science University College Cork 1 Outline Shortest Path Problem General Lemmas and Theorems. Algorithms Bellman-Ford
More informationChapter 6. Properties of Regular Languages
Chapter 6 Properties of Regular Languages Regular Sets and Languages Claim(1). The family of languages accepted by FSAs consists of precisely the regular sets over a given alphabet. Every regular set is
More informationThe Maximum Flow Problem with Disjunctive Constraints
The Maximum Flow Problem with Disjunctive Constraints Ulrich Pferschy Joachim Schauer Abstract We study the maximum flow problem subject to binary disjunctive constraints in a directed graph: A negative
More informationLecture 24: April 12
CS271 Randomness & Computation Spring 2018 Instructor: Alistair Sinclair Lecture 24: April 12 Disclaimer: These notes have not been subjected to the usual scrutiny accorded to formal publications. They
More informationGraphs and Network Flows IE411. Lecture 15. Dr. Ted Ralphs
Graphs and Network Flows IE411 Lecture 15 Dr. Ted Ralphs IE411 Lecture 15 1 Preflow-Push Algorithms First developed by A. V. Goldberg in 1985. Best preflow-push algorithms outperform best augmenting path
More informationIntroduction to Linear and Combinatorial Optimization (ADM I)
Introduction to Linear and Combinatorial Optimization (ADM I) Rolf Möhring based on the 20011/12 course by Martin Skutella TU Berlin WS 2013/14 1 General Remarks new flavor of ADM I introduce linear and
More informationDuality of LPs and Applications
Lecture 6 Duality of LPs and Applications Last lecture we introduced duality of linear programs. We saw how to form duals, and proved both the weak and strong duality theorems. In this lecture we will
More informationLecture notes for Analysis of Algorithms : Markov decision processes
Lecture notes for Analysis of Algorithms : Markov decision processes Lecturer: Thomas Dueholm Hansen June 6, 013 Abstract We give an introduction to infinite-horizon Markov decision processes (MDPs) with
More informationMaximum-Life Routing Schedule
Maximum-Life Routing Schedule Peng-Jun Wan wan@cs.iit.edu Peng-Jun Wan (wan@cs.iit.edu) Maximum-Life Routing Schedule 1 / 42 Outline Problem Description Min-Cost Routing Ellipsoid Algorithm Price-Directive
More informationFINAL EXAM PRACTICE PROBLEMS CMSC 451 (Spring 2016)
FINAL EXAM PRACTICE PROBLEMS CMSC 451 (Spring 2016) The final exam will be on Thursday, May 12, from 8:00 10:00 am, at our regular class location (CSI 2117). It will be closed-book and closed-notes, except
More informationCombinatorial Optimization
Combinatorial Optimization 2017-2018 1 Maximum matching on bipartite graphs Given a graph G = (V, E), find a maximum cardinal matching. 1.1 Direct algorithms Theorem 1.1 (Petersen, 1891) A matching M is
More informationLecture 2: Divide and conquer and Dynamic programming
Chapter 2 Lecture 2: Divide and conquer and Dynamic programming 2.1 Divide and Conquer Idea: - divide the problem into subproblems in linear time - solve subproblems recursively - combine the results in
More informationLecture 6 January 21, 2013
UBC CPSC 536N: Sparse Approximations Winter 03 Prof. Nick Harvey Lecture 6 January, 03 Scribe: Zachary Drudi In the previous lecture, we discussed max flow problems. Today, we consider the Travelling Salesman
More informationLecture 1. 1 Overview. 2 Maximum Flow. COMPSCI 532: Design and Analysis of Algorithms August 26, 2015
COMPSCI 532: Design and Analysis of Algorithms August 26, 205 Lecture Lecturer: Debmalya Panigrahi Scribe: Allen Xiao Oeriew In this lecture, we will define the maximum flow problem and discuss two algorithms
More informationAgenda. Soviet Rail Network, We ve done Greedy Method Divide and Conquer Dynamic Programming
Agenda We ve done Greedy Method Divide and Conquer Dynamic Programming Now Flow Networks, Max-flow Min-cut and Applications c Hung Q. Ngo (SUNY at Buffalo) CSE 531 Algorithm Analysis and Design 1 / 52
More informationMathematical Programs Linear Program (LP)
Mathematical Programs Linear Program (LP) Integer Program (IP) Can be efficiently solved e.g., by Ellipsoid Method Cannot be efficiently solved Cannot be efficiently solved assuming P NP Combinatorial
More informationCSCE423/823. Introduction. Flow Networks. Ford-Fulkerson Method. Edmonds-Karp Algorithm. Maximum Bipartite Matching 2/35 CSCE423/823.
1/35 2pt 0em Computer Science & Engineering 423/823 Design and s Lecture 07 (Chapter 26) Stephen Scott (Adapted from Vinodchandran N. Variyam) 2/35 Can use a directed graph as a flow network to model:
More information