SlideShare a Scribd company logo

Introduction to Approximation Algorithms

Jhoirene Clemente, profile picture
Jhoirene Clemente

The document discusses approximation algorithms for solving hard combinatorial optimization problems. It defines optimization problems and covers NP-hard problems like the clique, independent set, vertex cover, and traveling salesman problems. Approaches for solving NP-hard problems include exact algorithms, approximation algorithms that provide guaranteed good solutions, and heuristics without guarantees. Approximation algorithms aim to settle for good enough solutions rather than optimal ones.

Science
1 of 67
Downloaded 260 times
1
2
Most read
3
Most read
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
Approximation Algorithms Jhoirene B Clemente Algorithms and Complexity Lab Department of Computer Science University of the Philippines Diliman October 14, 2014
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Overview 1. Combinatorial Optimization Problems 2. NP-Hard Problems 2.1 Clique 2.2 Independent Set Problem 2.3 Vertex Cover 2.4 Traveling Salesman Problem 3. Approaches in Solving Hard Problems 4. Approximation Algorithms 5. Approximable Problems 6. Inapproximable Problems
50 Approximation Algorithms Jhoirene B Clemente 3 Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Optimization Problems [Papadimitriou and Steiglitz, 1998] Definition (Instance of an Optimization Problem) An instance of an optimization problem is a pair (F, c), where F is any set, the domain of feasible points; c is the cost function, a mapping c : F ! R The problem is to find an f 2 F for which c(f ) c(y) 8 y 2 F Such a point f is called a globally optimal solution to the given instance, or, when no confusion can arise, simply an optimal solution.
50 Approximation Algorithms Jhoirene B Clemente 3 Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Optimization Problems [Papadimitriou and Steiglitz, 1998] Definition (Instance of an Optimization Problem) An instance of an optimization problem is a pair (F, c), where F is any set, the domain of feasible points; c is the cost function, a mapping c : F ! R The problem is to find an f 2 F for which c(f ) c(y) 8 y 2 F Such a point f is called a globally optimal solution to the given instance, or, when no confusion can arise, simply an optimal solution. Definition (Optimization Problem) An optimization problem is a set of instances of an optimization problem.
50 Approximation Algorithms Jhoirene B Clemente 4 Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Optimization Problems [Papadimitriou and Steiglitz, 1998] Two categories 1. with continuous variables, where we look for a set of real numbers or a function 2. with discrete variables, which we call combinatorial, where we look for an object from a finite, or possibly countably infinite set, typically an integer, set, permutation, or graph.
50 Approximation Algorithms Jhoirene B Clemente 5 Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Combinatorial Optimization Problem Definition (Combinatorial Optimization Problem [Papadimitriou and Steiglitz, 1998]) An optimization problem = (D,R, cost, goal) consists of 1. A set of valid instances D. Let I 2 D, denote an input instance. 2. Each I 2 D has a set of feasible solutions, R(I ). 3. Objective function, cost, that assigns a nonnegative rational number to each pair (I , SOL), where I is an instance and SOL is a feasible solution to I. 4. Either minimization or maximization problem: goal 2 {min, max}.
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems 6 Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 NP-Hard Problems 1. Clique Problem 2. Independent Set Problem 3. Vertex Cover Problem 4. Traveling Salesman Problem
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems 7 Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Max Clique Problem Definition (Max Clique Problem) Given a complete weighted graph, find the largest clique
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems 7 Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Max Clique Problem Definition (Max Clique Problem) Given a complete weighted graph, find the largest clique Theorem Max Clique is NP-hard [Garey and Johnson, 1979].
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems 7 Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Max Clique Problem Definition (Max Clique Problem) Given a complete weighted graph, find the largest clique Theorem Max Clique is NP-hard [Garey and Johnson, 1979]. Proposition The decision variant of MAX-SAT is NP-Complete [Garey and Johnson, 1979].
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems 8 Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Max Clique Reduction from SAT Example (a _ b _ c) ^ (b _ ¯c _ ¯d) ^ (¯a _ c _ d) ^ (a _¯b _ ¯d)
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems 9 Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Max Clique Reduction from SAT
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique 10 Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Independent Set Definition (Max Independent Set Problem) Given a complete weighted graph, find the largest independent set
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique 10 Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Independent Set Definition (Max Independent Set Problem) Given a complete weighted graph, find the largest independent set Theorem Independent Set Problem is NP-hard [Garey and Johnson, 1979].
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem 11 Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Vertex Cover Problem Definition (Min Vertex Cover Problem) Given a complete weighted graph, find the minimum vertex cover.
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem 11 Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Vertex Cover Problem Definition (Min Vertex Cover Problem) Given a complete weighted graph, find the minimum vertex cover. Theorem Vertex Cover Problem is NP-hard [Garey and Johnson, 1979].
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem 12 Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Traveling Salesman Problem (TSP) Definition INPUT: Edge weighted Graph G = (V,E) OUTPUT: Minimum cost Hamiltonian Cycle.
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover 13 Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Approaches in Solving NP-hard Problems I Exact algorithm always obtain the optimal solution I Approximation algorithm settle for good enough solutions. Goodness of solution is guaranteed and measured using approximation ratio. I Heuristic algorithms produce solutions, which are not guaranteed to be close to the optimum. The performance of heuristics is often evaluated empirically.
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover 14 Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Approaches in Solving NP-hard Problems Design technique have a well specified structure that even provides a framework for possible implementations Concepts formulate ideas and rough frameworks about how to attack hard algorithmic problems .
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover 14 Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Approaches in Solving NP-hard Problems Design technique have a well specified structure that even provides a framework for possible implementations I Dynamic Programming I Greedy Schema I Divide and conquer I Branch and Bound I Local search I . . . Concepts formulate ideas and rough frameworks about how to attack hard algorithmic problems . I Approximation Algorithms I Parameterized Algorithms I Randomized Algorithms I . . .
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems 15 Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Approximation Algorithms Definition (Approximation Algorithm [Williamson and Shmoy, 2010] ) An -approximation algorithm for an optimization problem is a polynomial-time algorithm that for all instances of the problem produces a solution whose value is within a factor of of the value of an optimal solution. Given an problem instance I with an optimal solution Opt(I ), i.e. the cost function cost(Opt(I )) is minimum/maximum.
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems 15 Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Approximation Algorithms Definition (Approximation Algorithm [Williamson and Shmoy, 2010] ) An -approximation algorithm for an optimization problem is a polynomial-time algorithm that for all instances of the problem produces a solution whose value is within a factor of of the value of an optimal solution. Given an problem instance I with an optimal solution Opt(I ), i.e. the cost function cost(Opt(I )) is minimum/maximum. I An algorithm for a minimization problem is called -approximative algorithm for some 1, if the algorithm obtains a maximum cost of · cost(Opt(I )), for any input instance I .
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems 15 Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Approximation Algorithms Definition (Approximation Algorithm [Williamson and Shmoy, 2010] ) An -approximation algorithm for an optimization problem is a polynomial-time algorithm that for all instances of the problem produces a solution whose value is within a factor of of the value of an optimal solution. Given an problem instance I with an optimal solution Opt(I ), i.e. the cost function cost(Opt(I )) is minimum/maximum. I An algorithm for a minimization problem is called -approximative algorithm for some 1, if the algorithm obtains a maximum cost of · cost(Opt(I )), for any input instance I . I An algorithm for a maximization problem is called -approximative algorithm, for some 1, if the algorithm obtains a minimum cost of · cost(Opt(I )), for any input instance I .
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems 16 Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Approximation Ratio I Minimization, 1 cost(Opt(I )) cost(SOL) cost(Opt(I )) I Maximization, 1 cost(Opt(I )) cost(SOL) cost(Opt(I ))
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems 17 Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Approximation Ratio I Additive Approximation Algorithms: SOL OPT + c I Constant Approximation Algorithms: SOL c · OPT I Logarithmic Approximation Algorithms: SOL = O(log n) · OPT I Polynomial Approximation Algorithms: SOL = O(nc) · OPT, where c 1
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms 18 Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Example: Vertex Cover Problem Definition (Vertex Cover [Vazirani, 2001] ) Given a graph G = (V,E), a vertex cover is a subset C V such that every edge has at least one end point incident at C.
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms 18 Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Example: Vertex Cover Problem Definition (Vertex Cover [Vazirani, 2001] ) Given a graph G = (V,E), a vertex cover is a subset C V such that every edge has at least one end point incident at C.
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms 18 Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Example: Vertex Cover Problem Definition (Vertex Cover [Vazirani, 2001] ) Given a graph G = (V,E), a vertex cover is a subset C V such that every edge has at least one end point incident at C. Definition (Minimum Vertex Cover Problem) Given a complete weighted graph G = (V,E), find a minimum cardinality vertex cover C.
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms 19 Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Maximal Matching Definition (Matchings) Given a graph G = (V,E), a subset M E is called a matching if no two edges in M are adjacent in G. Figure : (a) Maximal matching (b) Perfect matching
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms 20 Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 2-Approximation Algorithm 1. Find a maximal matching in G, and output the set of matched vertices. Algorithm 1: 2-Approximation Algorithm for minimum vector cover problem
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms 21 Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 2-Approximation Algorithm Theorem Algorithm 1 is a 2-approximation algorithm.
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms 21 Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 2-Approximation Algorithm Theorem Algorithm 1 is a 2-approximation algorithm. Proof. Let M be a maximal matching in G = (V,E) and OPT be the minimum vertex cover in G, then |M| |OPT| The cover picked by the algorithm has cardinality |SOL| 2 · |M|. |SOL| 2 · |OPT|
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms 22 Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 2-Approximation Algorithm Example:
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms 22 Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 2-Approximation Algorithm Example:
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 23 Traveling Salesman Problem References CS 397 October 14, 2014 Solving Traveling Salesman Problem INPUT: Edge weighted graph 1. Compute a Minimum Spanning Tree for G 2. Select a vertex r 2 V(G) to be the root vertex 3. Let L be the list of vertices visited in a preorder walk OUTPUT: Hamiltonian Cycle that visits A preorder tree walk recursively visits every vertex in the tree, listing a vertex when it is first encountered , before any of its children are visited.
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 24 Traveling Salesman Problem References CS 397 October 14, 2014 Computing the Minimum Spanning Tree
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 25 Traveling Salesman Problem References CS 397 October 14, 2014 Computing the Minimum Spanning Tree
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 26 Traveling Salesman Problem References CS 397 October 14, 2014 Computing the Minimum Spanning Tree
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 27 Traveling Salesman Problem References CS 397 October 14, 2014 Computing the Minimum Spanning Tree
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 28 Traveling Salesman Problem References CS 397 October 14, 2014 Computing the Minimum Spanning Tree
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 29 Traveling Salesman Problem References CS 397 October 14, 2014 Computing the Minimum Spanning Tree
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 30 Traveling Salesman Problem References CS 397 October 14, 2014 Computing the Minimum Spanning Tree
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 31 Traveling Salesman Problem References CS 397 October 14, 2014 Computing the Minimum Spanning Tree
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 32 Traveling Salesman Problem References CS 397 October 14, 2014 Computing the Minimum Spanning Tree
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 33 Traveling Salesman Problem References CS 397 October 14, 2014 Computing the Minimum Spanning Tree
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 34 Traveling Salesman Problem References CS 397 October 14, 2014 Computing the Minimum Spanning Tree
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 35 Traveling Salesman Problem References CS 397 October 14, 2014 Preorder
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 36 Traveling Salesman Problem References CS 397 October 14, 2014 Preorder
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 37 Traveling Salesman Problem References CS 397 October 14, 2014 Preorder
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 38 Traveling Salesman Problem References CS 397 October 14, 2014 Preorder
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 39 Traveling Salesman Problem References CS 397 October 14, 2014 Preorder
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 40 Traveling Salesman Problem References CS 397 October 14, 2014 Preorder
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 41 Traveling Salesman Problem References CS 397 October 14, 2014 Preorder
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 42 Traveling Salesman Problem References CS 397 October 14, 2014 Hamiltonian Cycle from the Preorder Walk
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 43 Traveling Salesman Problem References CS 397 October 14, 2014 Hamiltonian Cycle from the Preorder Walk 4 + 4 + 10 + 15 + 10 + 16 = 59
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 44 Traveling Salesman Problem References CS 397 October 14, 2014 Total tour cost vs. the optimal cost Let H denote an optimal Hamiltonian tour in G. Let T be the minimum spanning tree of G. c(T) c(H) Let W be the full length cost of the preorder walk. c(W) = 2c(T) Let H be the output Hamiltonian Cycle from the algorithm c(W) c(H) Then c(H) 2c(H)
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 44 Traveling Salesman Problem References CS 397 October 14, 2014 Total tour cost vs. the optimal cost Let H denote an optimal Hamiltonian tour in G. Let T be the minimum spanning tree of G. c(T) c(H) Let W be the full length cost of the preorder walk. c(W) = 2c(T) Let H be the output Hamiltonian Cycle from the algorithm c(W) c(H) Then c(H) 2c(H) Approximation ratio () is 2
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 45 Traveling Salesman Problem References CS 397 October 14, 2014 Approximable Problems [Vazirani, 2001] Definition (APX) I An abbreviation for “Approximable, is the set of NP optimization problems that allow polynomial-time approximation algorithms with approximation ratio bounded by a constant.
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 45 Traveling Salesman Problem References CS 397 October 14, 2014 Approximable Problems [Vazirani, 2001] Definition (APX) I An abbreviation for “Approximable, is the set of NP optimization problems that allow polynomial-time approximation algorithms with approximation ratio bounded by a constant. I Problems in this class have efficient algorithms that can find an answer within some fixed percentage of the optimal answer.
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 46 Traveling Salesman Problem References CS 397 October 14, 2014 Polynomial-time Approximation Schemes Definition (PTAS [Aaronson et al., 2008]) The subclass of NPO problems that admit an approximation scheme in the following sense. For any 0, there is a polynomial-time algorithm that is guaranteed to find a solution whose cost is within a 1 + factor of the optimum cost. Contains FPTAS, and is contained in APX.
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 46 Traveling Salesman Problem References CS 397 October 14, 2014 Polynomial-time Approximation Schemes Definition (PTAS [Aaronson et al., 2008]) The subclass of NPO problems that admit an approximation scheme in the following sense. For any 0, there is a polynomial-time algorithm that is guaranteed to find a solution whose cost is within a 1 + factor of the optimum cost. Contains FPTAS, and is contained in APX. Definition (FPTAS [Aaronson et al., 2008] ) The subclass of NPO problems that admit an approximation scheme in the following sense. For any 0, there is an algorithm that is guaranteed to find a solution whose cost is within a 1 + factor of the optimum cost. Furthermore, the running time of the algorithm is polynomial in n (the size of the problem) and in 1/.
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 47 Traveling Salesman Problem References CS 397 October 14, 2014 Approximation Ratio of well known Hard Problems 1. FPTAS: Bin Packing Problem 2. PTAS: Makespan Scheduling Problem 3. APX: 3.1 Min Steiner Tree Problem (1.55 Approximable) [Robins,2005] 3.2 Min Metric TSP (3/2 Approximable) [Christofides,1977] 3.3 Max SAT (0.77 Approximable) [Asano,1997] 3.4 Vertex Cover (2 Approximable) [Vazirani,2001] 4. MAX SNP 4.1 Independent Set Problem 4.2 Clique Problem 4.3 Travelling Salesman Problem
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 48 Traveling Salesman Problem References CS 397 October 14, 2014 Complexity Classes [Ausiello et al., 2011] FPTAS ( PTAS ( APX ( NPO
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 49 Traveling Salesman Problem References CS 397 October 14, 2014 Inapproximable Problems Many problems have polynomial-time approximation schemes. However, there exists a class of problems that is not so easy [Williamson and Shmoy, 2010]. This class is called MAX SNP. Theorem For any MAXSNP-hard problem, there does not exist a polynomial-time approximation scheme, unless P = NP [Williamson and Shmoy, 2010].
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 49 Traveling Salesman Problem References CS 397 October 14, 2014 Inapproximable Problems Many problems have polynomial-time approximation schemes. However, there exists a class of problems that is not so easy [Williamson and Shmoy, 2010]. This class is called MAX SNP. Theorem For any MAXSNP-hard problem, there does not exist a polynomial-time approximation scheme, unless P = NP [Williamson and Shmoy, 2010].
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 49 Traveling Salesman Problem References CS 397 October 14, 2014 Inapproximable Problems Many problems have polynomial-time approximation schemes. However, there exists a class of problems that is not so easy [Williamson and Shmoy, 2010]. This class is called MAX SNP. Theorem For any MAXSNP-hard problem, there does not exist a polynomial-time approximation scheme, unless P = NP [Williamson and Shmoy, 2010]. Theorem If P6= NP, then for any constant 1, there is no polynomial-time approximation algorithm with approximation ratio for the general travelling salesman problem.
50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem 50 References CS 397 October 14, 2014 References I Aaronson, S., Kuperberg, G., and Granade, C. (2008). The complexity zoo.

More Related Content

PPTX
Presentation of daa on approximation algorithm and vertex cover problem
sumit gyawali
 
PDF
Approximation Algorithms
Nicolas Bettenburg
 
PPTX
NP completeness
Amrinder Arora
 
PPTX
Approximation algorithms
Bipesh Raj Subedi
 
PPTX
Brute force method
priyankabhansali217
 
PPT
Randomized algorithms ver 1.0
Dr. C.V. Suresh Babu
 
PPT
Np completeness
Rajendran
 
PPTX
Knowledge representation and Predicate logic
Amey Kerkar
 
Presentation of daa on approximation algorithm and vertex cover problem
sumit gyawali
 
Approximation Algorithms
Nicolas Bettenburg
 
NP completeness
Amrinder Arora
 
Approximation algorithms
Bipesh Raj Subedi
 
Brute force method
priyankabhansali217
 
Randomized algorithms ver 1.0
Dr. C.V. Suresh Babu
 
Np completeness
Rajendran
 
Knowledge representation and Predicate logic
Amey Kerkar
 

What's hot (20)

PPT
Sum of subsets problem by backtracking 
Hasanain Alshadoodee
 
PPTX
Vertex cover Problem
Gajanand Sharma
 
PPTX
Np completeness-Design and Analysis of Algorithms
adeel990
 
PDF
9. chapter 8 np hard and np complete problems
Jyotsna Suryadevara
 
PDF
All pairs shortest path algorithm
Srikrishnan Suresh
 
PDF
Undecidabality
Anshuman Biswal
 
PDF
Recurrent Neural Networks
Rakuten Group, Inc.
 
PDF
Greedy algorithm
International Islamic University
 
PPT
Genetic algorithm
garima931
 
PPTX
Genetic Algorithm
SEKHARREDDYAMBATI
 
PPTX
Tsp branch and-bound
Saravanan Natarajan
 
PPTX
A petri-net
Omar Al-Sabek
 
PDF
Design and analysis of algorithms
Dr Geetha Mohan
 
PDF
What is the Expectation Maximization (EM) Algorithm?
Kazuki Yoshida
 
PPTX
Dynamic programming
Yıldırım Tam
 
PPT
Chap8 new
Loc Tran
 
PDF
Understanding Bagging and Boosting
Mohit Rajput
 
PDF
P, NP, NP-Complete, and NP-Hard
Animesh Chaturvedi
 
PPTX
Computability and Complexity
Edward Blurock
 
PPT
Introduction to NP Completeness
Gene Moo Lee
 
Sum of subsets problem by backtracking 
Hasanain Alshadoodee
 
Vertex cover Problem
Gajanand Sharma
 
Np completeness-Design and Analysis of Algorithms
adeel990
 
9. chapter 8 np hard and np complete problems
Jyotsna Suryadevara
 
All pairs shortest path algorithm
Srikrishnan Suresh
 
Undecidabality
Anshuman Biswal
 
Recurrent Neural Networks
Rakuten Group, Inc.
 
Greedy algorithm
International Islamic University
 
Genetic algorithm
garima931
 
Genetic Algorithm
SEKHARREDDYAMBATI
 
Tsp branch and-bound
Saravanan Natarajan
 
A petri-net
Omar Al-Sabek
 
Design and analysis of algorithms
Dr Geetha Mohan
 
What is the Expectation Maximization (EM) Algorithm?
Kazuki Yoshida
 
Dynamic programming
Yıldırım Tam
 
Chap8 new
Loc Tran
 
Understanding Bagging and Boosting
Mohit Rajput
 
P, NP, NP-Complete, and NP-Hard
Animesh Chaturvedi
 
Computability and Complexity
Edward Blurock
 
Introduction to NP Completeness
Gene Moo Lee
 
Ad

Similar to Introduction to Approximation Algorithms (20)

PDF
Reoptimization techniques for solving hard problems
Jhoirene Clemente
 
PDF
DeepLearn2022 1. Goals & AlgorithmDesign.pdf
Sean Meyn
 
PDF
A Regularized Simplex Method
Gina Brown
 
PDF
Reoptimization Algorithms and Persistent Turing Machines
Jhoirene Clemente
 
PPT
Analysis of Algorithm
أحلام انصارى
 
PDF
Introduction to optimization Problems
Electronics & Communication Staff SCU Suez Canal University
 
PDF
Low Rank Optimisation, by Irène Waldspurger
Paris Women in Machine Learning and Data Science
 
PPT
Introduction to Optimization revised.ppt
JahnaviGautam
 
PDF
LP linear programming (summary) (5s)
Dionísio Carmo-Neto
 
PPT
Primal Dual
carlol
 
PPTX
Basic presentation of convex optimization 2018.pptx
ubaidqureshi27
 
DOCX
Mb0048 operations research
smumbahelp
 
PPTX
Approximation
RiyaSingh235
 
PDF
Fine Grained Complexity
AkankshaAgrawal55
 
PPT
Support Vector Machines
nextlib
 
PPT
Primal Dual
carlol
 
PPT
Approx
guest0264d3b
 
PDF
Molodtsov's Soft Set Theory and its Applications in Decision Making
inventionjournals
 
PDF
Molodtsov's Soft Set Theory and its Applications in Decision Making
inventionjournals
 
PDF
Boyd 4.6, 4.7
Koki Isokawa
 
Reoptimization techniques for solving hard problems
Jhoirene Clemente
 
DeepLearn2022 1. Goals & AlgorithmDesign.pdf
Sean Meyn
 
A Regularized Simplex Method
Gina Brown
 
Reoptimization Algorithms and Persistent Turing Machines
Jhoirene Clemente
 
Analysis of Algorithm
أحلام انصارى
 
Introduction to optimization Problems
Electronics & Communication Staff SCU Suez Canal University
 
Low Rank Optimisation, by Irène Waldspurger
Paris Women in Machine Learning and Data Science
 
Introduction to Optimization revised.ppt
JahnaviGautam
 
LP linear programming (summary) (5s)
Dionísio Carmo-Neto
 
Primal Dual
carlol
 
Basic presentation of convex optimization 2018.pptx
ubaidqureshi27
 
Mb0048 operations research
smumbahelp
 
Approximation
RiyaSingh235
 
Fine Grained Complexity
AkankshaAgrawal55
 
Support Vector Machines
nextlib
 
Primal Dual
carlol
 
Approx
guest0264d3b
 
Molodtsov's Soft Set Theory and its Applications in Decision Making
inventionjournals
 
Molodtsov's Soft Set Theory and its Applications in Decision Making
inventionjournals
 
Boyd 4.6, 4.7
Koki Isokawa
 
Ad

Recently uploaded (20)

PDF
HPLC-PPT.docx high performance liquid chromatography
darshanambiga1633
 
PDF
IFIT3 RNA-binding activity primores influenza A viruz infection and translati...
saracelre
 
PDF
bbec55_b34400a7914c42429908233dbd381773.pdf
agambirev
 
PPTX
LESSON 3_States of Matter and Particle Arrangement and Phase Changes.pptx
mariemoranteiron
 
PDF
Coordination Chemistry(Part-I) - Notes.pdf
divybahetravines
 
PDF
Agricultural Extension Presentation Slides
addisinicholas
 
PDF
NEET 2025 Megaa Solved Papers Collection
ebooksofficial2025
 
PDF
AlphaEarth Foundations and the Satellite Embedding dataset
gdgforscience
 
PPTX
Derivatives of integument scales, beaks, horns,.pptx
Dr Showkat Ahmad Wani
 
PPTX
neck nodes and dissection types and lymph nodes levels
DipuSarma
 
PPTX
1.pptx 2.pptx for biology endocrine system hum ppt
Dhananjaysankhla
 
PDF
Mastering Bioreactors and Media Sterilization: A Complete Guide to Sterile Fe...
Sabina Sharma
 
PPT
oscillatoria known as blue -green algae
Baher El-Nogoumy
 
PPTX
Animal Cell and plant cell for junior high school
Safira NurSa'adah
 
DOCX
The beginnings of Microbiology (discovery, development, scope, Identification...
Educator
 
PPTX
Comparative Structure of Integument in Vertebrates.pptx
Dr Showkat Ahmad Wani
 
PDF
N-enhancement in GN-z11: First evidence for supermassive stars nucleosynthesi...
Sérgio Sacani
 
PDF
VARICELLA VACCINATION: A POTENTIAL STRATEGY FOR PREVENTING MULTIPLE SCLEROSIS
ijab2
 
PPTX
ANEMIA WITH LEUKOPENIA MDS 07_25.pptx htggtftgt fredrctvg
vedvyas5668
 
PPTX
Excretory System in insects ( PPT Presentation)
Dr Showkat Ahmad Wani
 
HPLC-PPT.docx high performance liquid chromatography
darshanambiga1633
 
IFIT3 RNA-binding activity primores influenza A viruz infection and translati...
saracelre
 
bbec55_b34400a7914c42429908233dbd381773.pdf
agambirev
 
LESSON 3_States of Matter and Particle Arrangement and Phase Changes.pptx
mariemoranteiron
 
Coordination Chemistry(Part-I) - Notes.pdf
divybahetravines
 
Agricultural Extension Presentation Slides
addisinicholas
 
NEET 2025 Megaa Solved Papers Collection
ebooksofficial2025
 
AlphaEarth Foundations and the Satellite Embedding dataset
gdgforscience
 
Derivatives of integument scales, beaks, horns,.pptx
Dr Showkat Ahmad Wani
 
neck nodes and dissection types and lymph nodes levels
DipuSarma
 
1.pptx 2.pptx for biology endocrine system hum ppt
Dhananjaysankhla
 
Mastering Bioreactors and Media Sterilization: A Complete Guide to Sterile Fe...
Sabina Sharma
 
oscillatoria known as blue -green algae
Baher El-Nogoumy
 
Animal Cell and plant cell for junior high school
Safira NurSa'adah
 
The beginnings of Microbiology (discovery, development, scope, Identification...
Educator
 
Comparative Structure of Integument in Vertebrates.pptx
Dr Showkat Ahmad Wani
 
N-enhancement in GN-z11: First evidence for supermassive stars nucleosynthesi...
Sérgio Sacani
 
VARICELLA VACCINATION: A POTENTIAL STRATEGY FOR PREVENTING MULTIPLE SCLEROSIS
ijab2
 
ANEMIA WITH LEUKOPENIA MDS 07_25.pptx htggtftgt fredrctvg
vedvyas5668
 
Excretory System in insects ( PPT Presentation)
Dr Showkat Ahmad Wani
 

Introduction to Approximation Algorithms

  • 1. Approximation Algorithms Jhoirene B Clemente Algorithms and Complexity Lab Department of Computer Science University of the Philippines Diliman October 14, 2014
  • 2. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Overview 1. Combinatorial Optimization Problems 2. NP-Hard Problems 2.1 Clique 2.2 Independent Set Problem 2.3 Vertex Cover 2.4 Traveling Salesman Problem 3. Approaches in Solving Hard Problems 4. Approximation Algorithms 5. Approximable Problems 6. Inapproximable Problems
  • 3. 50 Approximation Algorithms Jhoirene B Clemente 3 Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Optimization Problems [Papadimitriou and Steiglitz, 1998] Definition (Instance of an Optimization Problem) An instance of an optimization problem is a pair (F, c), where F is any set, the domain of feasible points; c is the cost function, a mapping c : F ! R The problem is to find an f 2 F for which c(f ) c(y) 8 y 2 F Such a point f is called a globally optimal solution to the given instance, or, when no confusion can arise, simply an optimal solution.
  • 4. 50 Approximation Algorithms Jhoirene B Clemente 3 Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Optimization Problems [Papadimitriou and Steiglitz, 1998] Definition (Instance of an Optimization Problem) An instance of an optimization problem is a pair (F, c), where F is any set, the domain of feasible points; c is the cost function, a mapping c : F ! R The problem is to find an f 2 F for which c(f ) c(y) 8 y 2 F Such a point f is called a globally optimal solution to the given instance, or, when no confusion can arise, simply an optimal solution. Definition (Optimization Problem) An optimization problem is a set of instances of an optimization problem.
  • 5. 50 Approximation Algorithms Jhoirene B Clemente 4 Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Optimization Problems [Papadimitriou and Steiglitz, 1998] Two categories 1. with continuous variables, where we look for a set of real numbers or a function 2. with discrete variables, which we call combinatorial, where we look for an object from a finite, or possibly countably infinite set, typically an integer, set, permutation, or graph.
  • 6. 50 Approximation Algorithms Jhoirene B Clemente 5 Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Combinatorial Optimization Problem Definition (Combinatorial Optimization Problem [Papadimitriou and Steiglitz, 1998]) An optimization problem = (D,R, cost, goal) consists of 1. A set of valid instances D. Let I 2 D, denote an input instance. 2. Each I 2 D has a set of feasible solutions, R(I ). 3. Objective function, cost, that assigns a nonnegative rational number to each pair (I , SOL), where I is an instance and SOL is a feasible solution to I. 4. Either minimization or maximization problem: goal 2 {min, max}.
  • 7. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems 6 Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 NP-Hard Problems 1. Clique Problem 2. Independent Set Problem 3. Vertex Cover Problem 4. Traveling Salesman Problem
  • 8. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems 7 Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Max Clique Problem Definition (Max Clique Problem) Given a complete weighted graph, find the largest clique
  • 9. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems 7 Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Max Clique Problem Definition (Max Clique Problem) Given a complete weighted graph, find the largest clique Theorem Max Clique is NP-hard [Garey and Johnson, 1979].
  • 10. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems 7 Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Max Clique Problem Definition (Max Clique Problem) Given a complete weighted graph, find the largest clique Theorem Max Clique is NP-hard [Garey and Johnson, 1979]. Proposition The decision variant of MAX-SAT is NP-Complete [Garey and Johnson, 1979].
  • 11. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems 8 Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Max Clique Reduction from SAT Example (a _ b _ c) ^ (b _ ¯c _ ¯d) ^ (¯a _ c _ d) ^ (a _¯b _ ¯d)
  • 12. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems 9 Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Max Clique Reduction from SAT
  • 13. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique 10 Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Independent Set Definition (Max Independent Set Problem) Given a complete weighted graph, find the largest independent set
  • 14. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique 10 Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Independent Set Definition (Max Independent Set Problem) Given a complete weighted graph, find the largest independent set Theorem Independent Set Problem is NP-hard [Garey and Johnson, 1979].
  • 15. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem 11 Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Vertex Cover Problem Definition (Min Vertex Cover Problem) Given a complete weighted graph, find the minimum vertex cover.
  • 16. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem 11 Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Vertex Cover Problem Definition (Min Vertex Cover Problem) Given a complete weighted graph, find the minimum vertex cover. Theorem Vertex Cover Problem is NP-hard [Garey and Johnson, 1979].
  • 17. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem 12 Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Traveling Salesman Problem (TSP) Definition INPUT: Edge weighted Graph G = (V,E) OUTPUT: Minimum cost Hamiltonian Cycle.
  • 18. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover 13 Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Approaches in Solving NP-hard Problems I Exact algorithm always obtain the optimal solution I Approximation algorithm settle for good enough solutions. Goodness of solution is guaranteed and measured using approximation ratio. I Heuristic algorithms produce solutions, which are not guaranteed to be close to the optimum. The performance of heuristics is often evaluated empirically.
  • 19. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover 14 Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Approaches in Solving NP-hard Problems Design technique have a well specified structure that even provides a framework for possible implementations Concepts formulate ideas and rough frameworks about how to attack hard algorithmic problems .
  • 20. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover 14 Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Approaches in Solving NP-hard Problems Design technique have a well specified structure that even provides a framework for possible implementations I Dynamic Programming I Greedy Schema I Divide and conquer I Branch and Bound I Local search I . . . Concepts formulate ideas and rough frameworks about how to attack hard algorithmic problems . I Approximation Algorithms I Parameterized Algorithms I Randomized Algorithms I . . .
  • 21. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems 15 Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Approximation Algorithms Definition (Approximation Algorithm [Williamson and Shmoy, 2010] ) An -approximation algorithm for an optimization problem is a polynomial-time algorithm that for all instances of the problem produces a solution whose value is within a factor of of the value of an optimal solution. Given an problem instance I with an optimal solution Opt(I ), i.e. the cost function cost(Opt(I )) is minimum/maximum.
  • 22. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems 15 Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Approximation Algorithms Definition (Approximation Algorithm [Williamson and Shmoy, 2010] ) An -approximation algorithm for an optimization problem is a polynomial-time algorithm that for all instances of the problem produces a solution whose value is within a factor of of the value of an optimal solution. Given an problem instance I with an optimal solution Opt(I ), i.e. the cost function cost(Opt(I )) is minimum/maximum. I An algorithm for a minimization problem is called -approximative algorithm for some 1, if the algorithm obtains a maximum cost of · cost(Opt(I )), for any input instance I .
  • 23. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems 15 Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Approximation Algorithms Definition (Approximation Algorithm [Williamson and Shmoy, 2010] ) An -approximation algorithm for an optimization problem is a polynomial-time algorithm that for all instances of the problem produces a solution whose value is within a factor of of the value of an optimal solution. Given an problem instance I with an optimal solution Opt(I ), i.e. the cost function cost(Opt(I )) is minimum/maximum. I An algorithm for a minimization problem is called -approximative algorithm for some 1, if the algorithm obtains a maximum cost of · cost(Opt(I )), for any input instance I . I An algorithm for a maximization problem is called -approximative algorithm, for some 1, if the algorithm obtains a minimum cost of · cost(Opt(I )), for any input instance I .
  • 24. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems 16 Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Approximation Ratio I Minimization, 1 cost(Opt(I )) cost(SOL) cost(Opt(I )) I Maximization, 1 cost(Opt(I )) cost(SOL) cost(Opt(I ))
  • 25. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems 17 Approximation Algorithms Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Approximation Ratio I Additive Approximation Algorithms: SOL OPT + c I Constant Approximation Algorithms: SOL c · OPT I Logarithmic Approximation Algorithms: SOL = O(log n) · OPT I Polynomial Approximation Algorithms: SOL = O(nc) · OPT, where c 1
  • 26. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms 18 Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Example: Vertex Cover Problem Definition (Vertex Cover [Vazirani, 2001] ) Given a graph G = (V,E), a vertex cover is a subset C V such that every edge has at least one end point incident at C.
  • 27. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms 18 Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Example: Vertex Cover Problem Definition (Vertex Cover [Vazirani, 2001] ) Given a graph G = (V,E), a vertex cover is a subset C V such that every edge has at least one end point incident at C.
  • 28. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms 18 Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Example: Vertex Cover Problem Definition (Vertex Cover [Vazirani, 2001] ) Given a graph G = (V,E), a vertex cover is a subset C V such that every edge has at least one end point incident at C. Definition (Minimum Vertex Cover Problem) Given a complete weighted graph G = (V,E), find a minimum cardinality vertex cover C.
  • 29. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms 19 Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 Maximal Matching Definition (Matchings) Given a graph G = (V,E), a subset M E is called a matching if no two edges in M are adjacent in G. Figure : (a) Maximal matching (b) Perfect matching
  • 30. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms 20 Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 2-Approximation Algorithm 1. Find a maximal matching in G, and output the set of matched vertices. Algorithm 1: 2-Approximation Algorithm for minimum vector cover problem
  • 31. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms 21 Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 2-Approximation Algorithm Theorem Algorithm 1 is a 2-approximation algorithm.
  • 32. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms 21 Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 2-Approximation Algorithm Theorem Algorithm 1 is a 2-approximation algorithm. Proof. Let M be a maximal matching in G = (V,E) and OPT be the minimum vertex cover in G, then |M| |OPT| The cover picked by the algorithm has cardinality |SOL| 2 · |M|. |SOL| 2 · |OPT|
  • 33. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms 22 Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 2-Approximation Algorithm Example:
  • 34. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms 22 Vertex Cover Traveling Salesman Problem References CS 397 October 14, 2014 2-Approximation Algorithm Example:
  • 35. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 23 Traveling Salesman Problem References CS 397 October 14, 2014 Solving Traveling Salesman Problem INPUT: Edge weighted graph 1. Compute a Minimum Spanning Tree for G 2. Select a vertex r 2 V(G) to be the root vertex 3. Let L be the list of vertices visited in a preorder walk OUTPUT: Hamiltonian Cycle that visits A preorder tree walk recursively visits every vertex in the tree, listing a vertex when it is first encountered , before any of its children are visited.
  • 36. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 24 Traveling Salesman Problem References CS 397 October 14, 2014 Computing the Minimum Spanning Tree
  • 37. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 25 Traveling Salesman Problem References CS 397 October 14, 2014 Computing the Minimum Spanning Tree
  • 38. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 26 Traveling Salesman Problem References CS 397 October 14, 2014 Computing the Minimum Spanning Tree
  • 39. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 27 Traveling Salesman Problem References CS 397 October 14, 2014 Computing the Minimum Spanning Tree
  • 40. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 28 Traveling Salesman Problem References CS 397 October 14, 2014 Computing the Minimum Spanning Tree
  • 41. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 29 Traveling Salesman Problem References CS 397 October 14, 2014 Computing the Minimum Spanning Tree
  • 42. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 30 Traveling Salesman Problem References CS 397 October 14, 2014 Computing the Minimum Spanning Tree
  • 43. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 31 Traveling Salesman Problem References CS 397 October 14, 2014 Computing the Minimum Spanning Tree
  • 44. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 32 Traveling Salesman Problem References CS 397 October 14, 2014 Computing the Minimum Spanning Tree
  • 45. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 33 Traveling Salesman Problem References CS 397 October 14, 2014 Computing the Minimum Spanning Tree
  • 46. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 34 Traveling Salesman Problem References CS 397 October 14, 2014 Computing the Minimum Spanning Tree
  • 47. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 35 Traveling Salesman Problem References CS 397 October 14, 2014 Preorder
  • 48. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 36 Traveling Salesman Problem References CS 397 October 14, 2014 Preorder
  • 49. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 37 Traveling Salesman Problem References CS 397 October 14, 2014 Preorder
  • 50. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 38 Traveling Salesman Problem References CS 397 October 14, 2014 Preorder
  • 51. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 39 Traveling Salesman Problem References CS 397 October 14, 2014 Preorder
  • 52. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 40 Traveling Salesman Problem References CS 397 October 14, 2014 Preorder
  • 53. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 41 Traveling Salesman Problem References CS 397 October 14, 2014 Preorder
  • 54. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 42 Traveling Salesman Problem References CS 397 October 14, 2014 Hamiltonian Cycle from the Preorder Walk
  • 55. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 43 Traveling Salesman Problem References CS 397 October 14, 2014 Hamiltonian Cycle from the Preorder Walk 4 + 4 + 10 + 15 + 10 + 16 = 59
  • 56. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 44 Traveling Salesman Problem References CS 397 October 14, 2014 Total tour cost vs. the optimal cost Let H denote an optimal Hamiltonian tour in G. Let T be the minimum spanning tree of G. c(T) c(H) Let W be the full length cost of the preorder walk. c(W) = 2c(T) Let H be the output Hamiltonian Cycle from the algorithm c(W) c(H) Then c(H) 2c(H)
  • 57. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 44 Traveling Salesman Problem References CS 397 October 14, 2014 Total tour cost vs. the optimal cost Let H denote an optimal Hamiltonian tour in G. Let T be the minimum spanning tree of G. c(T) c(H) Let W be the full length cost of the preorder walk. c(W) = 2c(T) Let H be the output Hamiltonian Cycle from the algorithm c(W) c(H) Then c(H) 2c(H) Approximation ratio () is 2
  • 58. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 45 Traveling Salesman Problem References CS 397 October 14, 2014 Approximable Problems [Vazirani, 2001] Definition (APX) I An abbreviation for “Approximable, is the set of NP optimization problems that allow polynomial-time approximation algorithms with approximation ratio bounded by a constant.
  • 59. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 45 Traveling Salesman Problem References CS 397 October 14, 2014 Approximable Problems [Vazirani, 2001] Definition (APX) I An abbreviation for “Approximable, is the set of NP optimization problems that allow polynomial-time approximation algorithms with approximation ratio bounded by a constant. I Problems in this class have efficient algorithms that can find an answer within some fixed percentage of the optimal answer.
  • 60. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 46 Traveling Salesman Problem References CS 397 October 14, 2014 Polynomial-time Approximation Schemes Definition (PTAS [Aaronson et al., 2008]) The subclass of NPO problems that admit an approximation scheme in the following sense. For any 0, there is a polynomial-time algorithm that is guaranteed to find a solution whose cost is within a 1 + factor of the optimum cost. Contains FPTAS, and is contained in APX.
  • 61. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 46 Traveling Salesman Problem References CS 397 October 14, 2014 Polynomial-time Approximation Schemes Definition (PTAS [Aaronson et al., 2008]) The subclass of NPO problems that admit an approximation scheme in the following sense. For any 0, there is a polynomial-time algorithm that is guaranteed to find a solution whose cost is within a 1 + factor of the optimum cost. Contains FPTAS, and is contained in APX. Definition (FPTAS [Aaronson et al., 2008] ) The subclass of NPO problems that admit an approximation scheme in the following sense. For any 0, there is an algorithm that is guaranteed to find a solution whose cost is within a 1 + factor of the optimum cost. Furthermore, the running time of the algorithm is polynomial in n (the size of the problem) and in 1/.
  • 62. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 47 Traveling Salesman Problem References CS 397 October 14, 2014 Approximation Ratio of well known Hard Problems 1. FPTAS: Bin Packing Problem 2. PTAS: Makespan Scheduling Problem 3. APX: 3.1 Min Steiner Tree Problem (1.55 Approximable) [Robins,2005] 3.2 Min Metric TSP (3/2 Approximable) [Christofides,1977] 3.3 Max SAT (0.77 Approximable) [Asano,1997] 3.4 Vertex Cover (2 Approximable) [Vazirani,2001] 4. MAX SNP 4.1 Independent Set Problem 4.2 Clique Problem 4.3 Travelling Salesman Problem
  • 63. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 48 Traveling Salesman Problem References CS 397 October 14, 2014 Complexity Classes [Ausiello et al., 2011] FPTAS ( PTAS ( APX ( NPO
  • 64. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 49 Traveling Salesman Problem References CS 397 October 14, 2014 Inapproximable Problems Many problems have polynomial-time approximation schemes. However, there exists a class of problems that is not so easy [Williamson and Shmoy, 2010]. This class is called MAX SNP. Theorem For any MAXSNP-hard problem, there does not exist a polynomial-time approximation scheme, unless P = NP [Williamson and Shmoy, 2010].
  • 65. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 49 Traveling Salesman Problem References CS 397 October 14, 2014 Inapproximable Problems Many problems have polynomial-time approximation schemes. However, there exists a class of problems that is not so easy [Williamson and Shmoy, 2010]. This class is called MAX SNP. Theorem For any MAXSNP-hard problem, there does not exist a polynomial-time approximation scheme, unless P = NP [Williamson and Shmoy, 2010].
  • 66. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover 49 Traveling Salesman Problem References CS 397 October 14, 2014 Inapproximable Problems Many problems have polynomial-time approximation schemes. However, there exists a class of problems that is not so easy [Williamson and Shmoy, 2010]. This class is called MAX SNP. Theorem For any MAXSNP-hard problem, there does not exist a polynomial-time approximation scheme, unless P = NP [Williamson and Shmoy, 2010]. Theorem If P6= NP, then for any constant 1, there is no polynomial-time approximation algorithm with approximation ratio for the general travelling salesman problem.
  • 67. 50 Approximation Algorithms Jhoirene B Clemente Optimization Problems Hard Combinatorial Optimization Problems Clique Independent Set Problem Vertex Cover Approaches in Solving Hard Problems Approximation Algorithms Vertex Cover Traveling Salesman Problem 50 References CS 397 October 14, 2014 References I Aaronson, S., Kuperberg, G., and Granade, C. (2008). The complexity zoo.