Fundamentals of Algorithms (cs502)

Assignment No.5

Total Marks: 20
Due Date: 13th of July, 2010

Objectives
Objective of this assignment is to make students able to understand and implement
Dynamic Programming, Greedy Algorithms and basic concepts relating to Graphs.
The topics covered in this assignment are,

a. Coin Change Problem

Important Instructions
a. More than one correct solution of a single question is possible
b. Please do your own work, avoid copying solutions of questions from internet or
any other resource.

Question No.1
Coin Change Problem:
Suppose a country ABC’s currency has the following coins,

a. 3 cent
b. 6 cent
c. 9 cent

1. You have to find the solution of coin change problem using Dynamic
Programming approach for amount 12. [Your solution should have
minimum number of coins].
2. Find the solution of the above problem using Greedy Algorithm’s approach.

Hints:
o Your Table will have coins (denominations) as rows and values from 1 to
12 as columns. This problem is very similar to knapsack problem with the
difference that here value is count of coins and we will select minimum
coins count, you can see further details in handouts and video lecture.
o Since we want to minimize the number of coins used,
C[i, j] = min(C[i - 1, j], 1 + C[i, j - di])
o We will find the coins to be used in similar manner using separate table as
we found in Knapsack problem.
o For Greedy Approach you will take coin of highest value first.
Question No.2
Huffman Encoding:

1. Suppose our text include only characters A, B, C, D, E and f and you have to
encode the following string,

“AAABBDDDDEEEBABAABBB”

Encode this string using Huffman Encoding and showing all steps including
finding probability of each character, making Optimal Huffman Encoding Tree
and finding codes for each character. Also show the percentage compression
achieved using Huffman Encoding for this string if original string was stored in
the form of ASCII characters (8 bits for each character).

2. We can construct Huffman Encoding Tree using many ways; describe

Hints:
o Probability of each character will be equal to count of occurrence of that
character divided by total characters occurrence count.
o Consult pseudo code for constructing Huffman Encoding Tree described
in lecture to find optimal way of constructing Huffman Encoding Tree.

Question No.3
Activity Selection Problem:

While studying Activity Selection Problem we realized that sometimes we are able to
find some heuristic (greedy algorithm) that help us to find one of the optimal solutions
very efficiently. The heuristic was,

“We sorted the given activities on the base of their final times, using the fact that all
activities in an acceptable schedule should have start time greater than or equal to finish
time of previous activity. Then we scanned through all activities and keep on adding
activities having start time greater than or equal to finish time of previous activity (note
that first activity will always be in an acceptable schedule and our greedy algorithm will
always result in same optimal solution (set of compatible activities)”

Consider the following activities table sorted on the base of their finish times

i 1 2 3 4 5 6 7 8
si 1 3 2 4 3 5 6 8
fi 3 4 6 8 8 9 10 11

1. Find the optimal schedule using Dynamic Programming.

o We will add two pseudo activities a0 and a9 on both sides of these eighth
activities because,

We need to compare the starting time of a1 with finish time of previous activity
a0, similarly we need to compare the final time of last activity with starting time
of its next activity. So we take finish time of a0 as f0 = 0 and start time of a9 as s9
= infinite, as shown in table below

i 0 1 2 3 4 5 6 7 8 9
si 1 3 2 4 3 5 6 8 infinite
fi 0 3 4 6 8 8 9 10 11

The Dynamic Programming Table structure for this problem will as follows,
i
0 1 2 3 4 5 6 7 8 9
0
1
2
3
j 4
5
6
7
8
9

Only shaded cells will be filled as for other cells i >= j ( As activities are sorted on the
base of finish time so higher number activities can not come before lower number
activities hence un-shaded cells will be zero as they can never come in compatible
schedule).
Cells will be filled as follows,
For any cell [i , j] , we will write the count of activities that are compatible with those
activities ( lie between time interval [f i , s j], i.e they start after activity i finishes
and finish before activity j starts.

The formula to calculate any cell value will be,

⎧ 0 ifS [i, j ] = φ ⎫
C[i, j ] = ⎨ ⎬
⎩C[i, j] = max [C[i, k] + C[k, j] + 1] i<k< j fS [i, j ] ≠ φ ⎭
 It means we will have to look the activities table to find that any activity between i
and j is compatible with them if so we will write value according to above formula
and we will write activity no. in k table.
o The diagonal cell entries will be zero as for these cells i = j
o For cells corresponding to two consecutive activities like C[0,1], C[1,2], C[2,3], and
so on… We will have Sij = φ , from this hint you can easily judge their values from
above equation.
o After these cells as we will move towards top right corner we will have one or more
activities between i and j and we will check whether any of them is compatible with
activity I and j and we will find values for C and K table.
For cell C [0,2]
Our k can be 1 only [i< k < j]
So see whether activity 1 is compatible with 0 and 2, answer is Yes, as from table of
activities
a0 Æ a1 Æ a2
is acceptable as,
f0 <= s1<f1 <= s2 < f2
0 <= 1 < 3 <= 3 < 4

In simple words a1 is compatible with a0 and a2 [can be run between a1 and a2]
C[0,2] = C[0,1]+C[1,2]+1 = 0 + 0 + 1 = 1
K= 1
C[0,2] value will be filled in C table and k value will be filled in K table.
For cell C [0,3]
Our k can be 1 or 2 [i< k < j]
So first see whether activity 1 is compatible with 0 and 3, answer is No as 3 start time
is 2 and 1 finish time is 3 so they can not be compatible. Now check same for activity
2, answer is again No, so
Sij = φ , and above row of formula will be satisfied.

Question No.4
Graphs:
Give the answer of following questions:

a. Difference between a Tree and a Graph ?

Assignment Uploading Instructions: