This preview shows pages 1–4. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: CIS 502  Spring 2005 Final and WPE Exam: 5/03/05 1. (20 points) (a) Prove that if there is a fully polynomialtime approximation scheme for vertex cover, then P = NP. Answer: In a n vertex graph, the size of a minimum vertex cover is between 0 and n . If there is a fully polynomial approximation scheme for vertex cover, we can choose = 1 / ( n + 1) and get an algorithm that produces a vertex cover of size at most OPT (1+ 1 n +1 ). Since OPT is the largest integer less than or equal to this size, the algorithm must in fact produce an optimal vertex cover. Furthermore its running time is polynomial in n and 1 / = ( n + 1), which is polynomial in the input size. Thus, we would have a polynomial time algorithm for an NPhard problem which shows that P = NP. (b) Let G be an arbitrary graph. Let M 1 and M 2 be any two maximal matchings in G . Prove that  M 1  ≥  M 2  2 . (Recall that the size of a matching M is denoted  M  and is equal to the number of edges in M .) Answer: There are many good proofs for this problem. One that I read in several of your papers is as follows: Suppose for contradiction that M 1 and M 2 are maximal matchings with  M 1  <  M 2  / 2. For each edge ( u,v ) in M 2 at least one of the endpoints must be matched by M 1 . Otherwise, we could have added ( u,v ) to M 1 contradicting the assumption that it is maximal. Each edge in M 1 covers two vertices and hence the  M 1  edges can only hit 2  M 1  <  M 2  of the pairs of vertices matched by M 2 . Another elegant proof from one of you: Let k be the size of the optimal vertex cover. From class we know that  M 1  ≤ k ≤ 2  M 1  M 2  ≤ k ≤ 2  M 2  Combining the first inequality from the first line and the second from the second we get  M 1  ≤ 2  M 2  and symmetrically  M 2  ≤ 2  M 1 . 1 2. (20 points) The problem Half Clique is defined below. Problem: Half Clique Instance: Graph G = ( V,E ) with an even number of nodes....
View Full
Document
 Spring '09
 Naver

Click to edit the document details