National University of Computer & Emerging Sciences, Karachi

Fall-2023 Department of Computer Science

Final Exam
18th December 2023, 09:30 AM – 12:00 PM
Part (B)
Course Code: CS2009 Course Name: Design and Analysis of Algorithm
Instructor Name / Names: Dr. Muhammad Atif Tahir, Dr. Fahad Sherwani, Dr. Farrukh
Saleem
Student Roll No: Section:

Instructions:
• Return the question paper.
• Read each question completely before answering it. There are 8 questions on 4 pages.
• In case of any ambiguity, you may make assumptions. But your assumption should not contradict
any statement in the question paper.

Time: 150 minutes Max Marks: 40

Question # 3 [CLO:1] Answer the following questions and explain in only 2-3 lines only. [5
marks]

Strict Marking for T/F. Rest can be partial depends upon the correct answer

a) [0.5 Point] Floyd-Warshall algorithm computes shortest path from every vertex to every other vertex
using O(n3) complexity, whereas Bellman-Ford can compute in O(n3), only from a single source to
other vertices. Why one needs to use Bellman-Ford algorithm, in presence of more information rich
Floyd-Warshall algorithm. (apart from detecting a negative cycle)
b) [0.5 Points] True / False: Does Directed Acyclic Graph (DAG) can have negative weight edges?
c) [0.5 Point] Describe the difference between tight and loose bounds in Big-O notation. Provide an
example scenario where a tight bound is preferable and explain why.
d) [0.5 Point] What do you mean by the optimal solution?
e) [1 Point] Problem A reduces to problem B if you can use an algorithm that solves B to help solve A.
Let’s say problem A is to find median and problem B uses any sorting method with O(n2) complexity.
What would be the complexity of Problem A
f) [0.5 Point] True or False: In dynamic programming, a subproblem do share sub-subproblems

g) [1.5 Points] Differentiate b/w Single-source, Single-pair and All-pairs shortest path and one algorithm
for each studied during this course

Question # 4 [CLO: 3] [5 marks]

Demand for multimedia, combining audio, video, and data streams over a network, is rapidly
increasing. Some of the most popular uses of multimedia are real-time interactive applications
such as desktop video and audio conferencing, collaborative engineering, shared white boards,
transmission of university lectures to a remote audience, and animated simulations. With the
advent of real-time interactive applications, delay constraint is also an important objective along
with minimizing bandwidth cost. Delay constraint means that a packet must be reached within
that interval. Figure below shows an example tree. Each edge consists of 2 values (first one is
bandwidth cost and second one is delay). Your objective is to design a spanning tree that
optimizes both bandwidth cost and delay constraint simultaneously. Later show the dry run using
Figure below. [Hint: Modify Prims or Kruskal using some weighted function].

b c
(2,1)

(10,2) 5,1
(1,3)

d e h
(7,1) (8,1)
a (3,2)

(2,1) (9,1)
(1,3) (7,1)

(10,1)
g f

Sol:

Ideally, 1st normalize both bandwidth cost and delay constraint to give equal importance to both.

Then apply following.

Modify Cost function = 0.5 * bandwidth + 0.5 * delay constraint (Ideally first normalize both
parameters using max-min or other normalize function but for simplicity the above solution is
partially acceptable)

b-c = 0.5 * 2 + 0.5 * 1 = 1.5; Similarly for others and then apply either Kruskal or Prims

Marking Scheme: Correct Modification of normalized cost function = 2 Points.

Correct Modification without normalized cost function = 1 Point
Running of Kruskal or Prims = 3 marks showing steps correctly. 3 marks to be given if
they just attempt to run prims or Kruskal independently.
Question # 5 [CLO: 3] [5 marks]

Longest palindrome subsequence

A palindrome is a nonempty string over some alphabet that reads the same forward and backward.
Examples of palindromes are all strings of length 1, civic, racecar etc. Give an efficient algorithm
to find the longest palindrome that is a subsequence of a given input string. For example, given
the input “character”, your algorithm should return “carac”.
The idea is to check and compare the first and last characters of the string. There are only two
possibilities for the same: If the first and the last characters are the same, it can be ensured that
both the characters can be considered in the final palindrome and hence, add 2 to the result, since
we have found a sequence of length 2 and recurse the remaining substring S[i + 1, j – 1]. In case,
the first and the last characters aren’t the same, the following operation must be performed:
Recurse S[i + 1, j]
Recurse S[ i, j – 1]
a) Design an algorithm using a brute force approach (2^n possible subsequences) to find the
maximum length obtained amongst them [2 Points]
b) Design an algorithm using dynamic programming approach (polynomial time solution
required) to find the maximum length obtained amongst them [3 Points]

For both parts (a) and (b), only pseudocode is required but must be with clear formulas / loops/ if-
else statements not just plain text. You can show dry run for your own understanding, but dry run
would not have any weightage.

Solution:
Brute Force Algorithm
LPS(1..n) {
if (n == 1)
return 1
if (S[1] == S[n])
return 2 + LPS(2..n-1)
else
return max( LPS(1..n-1), LPS(2..n) )
}

Marking Scheme: Correct Sol 2 Points; Partial: 0.5-1.5 Points, No Solution: 0

The Dynamic Programming Algorithm

int LPS[1..n][1..n];

DPLPS(char S[1..n]) {
for (i = 1; i <= n; i++)
LPS[i][i] = 1;
for (len = 1; len < n; len++) {
 for (i = 1; i < n-len; i++) {
j = i + len;
if (S[j] == S[j])
LPS[i][j] = 2 + LPS[i+1[j-1];
else
LPS[i][j] = max{LPS[i][j-1], LPS[i+1][j]}
}
}
return LPS[1][n];
}

Marking Scheme: Correct Sol 3 Points; Partial: 0.5-2.5 Points, No Solution: 0

Question # 6 [CLO: 1] [5 marks]

Given an array of integers arr[] of size N and an integer, the task is to rotate the array elements to
the left by d positions, (where d<=n).
a) Design an algorithm to solve the above task with O(n) time complexity, and O(n) space
complexity [2.5 Points].
b) Design an algorithm to solve the above task with O(n^2) time complexity, and O(1) space
complexity [2.5 Points]

Pseudocode in plain text is acceptable for this question but must be clear.

Example:
Input: arr[] = {3, 4, 5, 6, 7, 1, 2}, d=2
Output: 5 6 7 1 2 3 4
Solution:
Algorithm a (Using temp array):
• First store the elements from index d to N-1 into the temp array.
• Then store the first d elements of the original array into the temp array.
• Copy back the elements of the temp array into the original array

Marking Scheme: Correct Sol 2.5 Points; Partial: 0.5-2.0 Points, No Solution: 0

Algorithm b (Rotate one by one):

• At each iteration, shift the elements by one position to the left circularly (i.e., first element
becomes the last).
• Perform this operation d times to rotate the elements to the left by d position.

Marking Scheme: Correct Sol 2.5 Points; Partial: 0.5-2.0 Points, No Solution: 0

Question # 7 [CLO:3] [5 marks]

Calculate the Levenshtein-Distance for str1 = “PLASMA”, and str2 = “ALTRUISM” using
Dynamic Programming Approach of Levenshtein-Distance.
Pseudocode Levenshtein-Distance is given below.
Levenshtein-Distance is the minimum edit distance between two strings is the minimum number
of editing operations:
- Insertion
- Deletion
- Substitution
Needed to transform one into the other. Examples include Spell correction, computational
biology etc. Normally, each operation has cost of 1. However, in Levenshtein, the substitution
cost can be taken as 2.

For two strings X of length n Y of length m. We define D(i,j) the edit distance between X[1..i]
and Y[1..j] i.e., the first i characters of X and the first j characters of Y. The edit distance
between X and Y is thus D(n,m).
Algorithm
//Initialization D(i,0) = i
D(0,j) = j
//Recurrence Relation:
For each i = 1…n
For each j = 1…m
D(i,j)= min(
D(i-1, j) + 1,
D(i, j-1) + 1,
D(i-1, j-1) + 2, if X(i) ≠ Y(j)
D(i-1, j-1) + 0 if X(i) = Y(j) )
//Termination: D(n,m) is distance

Sol:

Marking Scheme: For each error; deduct: 0.5 marks

Question # 8 [CLO: 5] [5 marks]

While visiting the Northern Areas of Pakistan you asked your travel guide to provide you
information regarding the famous tourist points. Your travel guide handed you the following table
showing four tourist points and the distances between these locations. You are determined to visit
all tourist points once and return to the starting location besides reducing your travel expenses.
Based on your knowledge in this course how would travel to these locations. Show all the
computations and steps to solve it

𝐴1 𝐴2 𝐴3 𝐴4

𝐴1 0 2 1 4
 𝐴2 2 0 3 2

𝐴3 1 3 0 5

𝐴4 4 2 5 0

Marking Scheme: Correct Naming of Solution i.e. TSP: 2 Points followed by marks on correct
steps and solution b/w 0 -3

Any Greedy Solution e.g. Steiner Tree etc without within 2 times of optimal solution = 2 Points
(Maximum)
Question # 9 [CLO: 5] [5 marks]

Let φ = (x1 v x1 v x2) Λ (¬x1 v ¬x2 v ¬x2) Λ (¬x1 v x2 v x2) which is 3-CNF formula with m=2
variables and L=3 clauses.
a) Find the value of k = m + 2L [0.25 Points]
b) Prove that Φ is satisfiable iff GΦ has a vertex cover of size k [3.75 Points]
[Hint: You can directly think of a solution, or two step solution i) reduce 3SAT to
Independent Set ii) Reduce Independent Set to k-Vertex Cover]
c) From k-vertex cover solution above, find whether x1x2 = {00, 01, 10, 11} is satisfiable or
not [1 Point]
Solution:

Marking Scheme

a) Marking Scheme: Correct Sol 0.5 Points; Any other / no solution: 0

b) Correct Sol 3.75 Points; Partial: 0.5-3.5 Points, No Solution: 0
c) Correct Sol for x1x2 0.25 Points each
Question # 10 [CLO: 3] [5 marks]

Apply the Bellman-Ford shortest path algorithm on the following graph, starting from vertex "a".
Show all steps

Solution

4 Points: There is negative cycle. Students should complete and optionally show 2nd loop of the
algorithm. -0.5 for each wrong entry
# edges A b C d e
0 0 ∞ ∞ ∞ ∞
1 0 ∞ ∞ -2 5
2 0 8 0 -2 -5
3 0 -2 -10 -2 -5
4 -1 -2 -14 -2 -5

1 Point for this table or showing of correct Prev [ vertex]

Vertex a b c d e
Prev[vertex] b e b a d