Name : NALEKAR MUKESH SHANTARAM PREMA

BITS ID NO. : 2022da04307

Course Name : Data Structures and Algorithms Design

A)

Greedy Algorithm,

1. Start with an empty set S.

2. Repeat until no vertices are left in the graph: a. Choose a vertex v with minimum degree in
the remaining graph. b. Add v to S. c. Delete v and all its neighbors from the graph.

3. Return S as the maximal independent set.

Proof of Correctness: Assume that the algorithm outputs S as the maximal independent set for a
given graph G. We need to prove that S is indeed a maximal independent set, i.e., we cannot add any
more vertices to S without violating the independence condition.

Let us assume that there exists a vertex u in G that is not in S but can be added to S without violating
the independence condition. Since u is not in S, it must have been deleted from G along with all its
neighbors during the execution of the algorithm. Thus, all the neighbors of u are not in G, which
means that they have already been deleted from G. Therefore, u must have had a degree greater
than the degree of v (the vertex with minimum degree in the remaining graph) when the algorithm
selected v.

This contradicts the choice of v, as v was chosen as the vertex with minimum degree in the
remaining graph. Hence, the assumption that there exists a vertex u in G that is not in S but can be
added to S without violating the independence condition is false. Therefore, S is indeed a maximal
independent set for the given graph G.

Example: Consider the following graph:

1 -- 2 -- 3

| | |

4 -- 5 -- 6

If we apply the Maximum Independent Set (G) algorithm on this graph, we get the following steps:

1. S = {} (empty set)

2. Choose vertex 1 (minimum degree = 1), S = {1} Delete vertices 1 and 4 (neighbors of 1)

3. Choose vertex 2 (minimum degree = 2), S = {1, 2} Delete vertices 2, 5, 3 (neighbors of 2)

4. S = {1, 2} is the maximal independent set.

B)

In the case of Dijkstra's algorithm for finding the single source shortest path in a graph, the choice of
representation of the graph can indeed have an impact on the efficiency of the algorithm.

Aruna is correct in stating that the best way to represent this graph for Dijkstra's algorithm is to use
an adjacency list. This is because an adjacency list can efficiently represent sparse graphs like the
one given in the question. In an adjacency list representation, each vertex is represented as a node
in a linked list, and the list contains all the adjacent vertices along with their weights. For the given
graph, since most of the vertices have only one or two adjacent vertices, using an adjacency list
would result in a more efficient representation than using an adjacency matrix.

Bablo is incorrect in stating that the best way to represent the graph is using an adjacency matrix. An
adjacency matrix can efficiently represent dense graphs, where most vertices have many adjacent
vertices, but for sparse graphs like the one given in the question, the matrix representation would
contain many entries with a value of zero, which would be wasteful in terms of space.

Cathy's statement that the choice of representation does not matter is incorrect. While the same
Dijkstra's algorithm can be applied to both representations, the time complexity of the algorithm can
be affected by the choice of representation. In the case of sparse graphs, as in the given example, an
adjacency list representation would be more efficient than an adjacency matrix representation.

Therefore, Aruna is correct in stating that the best way to represent this graph is to use an adjacency
list.

C)

If problem G can be reduced to Clique via a deterministic polynomial time reduction and Clique is an
NP-complete problem, then it follows that problem G is at least as hard as NP-complete. Therefore,
problem G is in the NP-complete complexity class, or in other words, G is an NP-complete problem
itself.

To be more specific, since G can be reduced to Clique, which is an NP-complete problem, we can
conclude that G is at least as hard as Clique, and since Clique is NP-complete, G must also be NP-
complete.