SWEG 5012

ADVANCED ALGORITHMS
AND PROBLEM SOLVING

LECTURE 1 INTRODUCTION
BEAKAL GIZACHEW

Based on slides of Serdar Taşıran and PiyushKumar and

adapted by Alptekin Küpçü
 ALGORITHMS
• Algorithm: Named after the 9th century mathematician el-Harezmi (al-
Khwarizmi)
• An algorithm: A tool for solving a well-specified computational problem.
• Problem statement:
• Inputs, outputs
• The desired input/output relationship
• Algorithm describes a specific computational procedure for producing
the required output
• Definiteness : Each step should be defined precisely
• Correctness : Should produce correct output for each input
• Finiteness : Should produce desired output after a finite number of steps
• Effectiveness: Each step should be performed exactly in a finite amount of time
• Generality: Should be applicable to any case (not just particular inputs)

We are interested in designing algorithms and

analyzing their performance in this course.
2
EFFICIENCY
• Hardware and memory are fast and cheap. Computing
getting cheaper and cheaper. Intelligent manpower is
expensive
• Assumptions:
• Single processor, random-access machine
• Algorithm steps are executed sequentially
• Each steps take 1 unit of time
• Makes our results machine-independent
• Otherwise impossible to compare!
• I also assume you already know:
• Big-Oh notation and asymptotic complexity analysis
• Basic data structures and algorithms
• How to analyze recurrence relations

3
ANALYZING ALGORITHMS
Is the algorithm correct?
• Does it terminate on all inputs?
• Does it produce the required output?
What amount of resources does the algorithm use up?
• Memory
• Communication bandwidth
• Number of logic gates (if implemented in hardware), speed (depth) of
logic circuit
• Running time

4
EFFICIENT ALGORITHMS
A bad approach uses up computing power faster than you might
think. Examples:
• Sorting a million numbers (emails, files, etc.)

O(n2) 2 n2 109 2000 sec.

algorithm instructions inst/second
O(n log n) 50 n log n 109 1 sec.
algorithm instructions inst/second

• Interactive graphics: Algorithms must terminate in 1/30 of a sec.

5
OUR FIRST ALGORITHM:
STABLE MARRIAGE

6
THE PROBLEM
• Input: There are n men and n women
• Each man has a preference list, so does each woman.
• These lists have no ties.
• Devise a system by which each of the n men and n women can end
up getting married.

• Different metrics possible:

• Maximize the number of people who get their first match?
• Maximize the average satisfaction?
• Maximize the minimum satisfaction?
• Can anything go wrong?

7
SAMPLE PREFERENCE LISTS

st nd rd st nd rd
Man 1 2 3 Woman 1 2 3
X A B C A Y X Z
Y B A C B X Y Z
Z A B C C X Y Z

unstable pair: man m and women w are unstable if

• m prefers w to his assigned match
• w prefers m to his assigned match
What is wrong?
Unstable pairs: (X,C) and (B,Y)
X and B prefer each other to their current pairs.
8
STABLE MATCHING

st nd rd st nd rd
Man 1 2 3 Woman 1 2 3
X A B C A Y X Z
Y B A C B X Y Z
Z A B C C X Y Z

Anything wrong?
No pair creating instability.

9
ANOTHER STABLE MATCHING

st nd rd st nd rd
Man 1 2 3 Woman 1 2 3
X A B C A Y X Z
Y B A C B X Y Z
Z A B C C X Y Z

10
STABILITY IS PRIMARY
• Any reasonable list of criteria must contain the stability
criterion.
• A pairing is doomed if it contains a rogue couple.

• Solution Idea: Allow the pairs to keep breaking up and

reforming until they become stable

• Can we argue that the couples will not continue breaking up

and reforming forever?

11
Review Question
• Brute force algorithm – an algorithm that checks every
possible solution
• In terms of n, what is the running time for brute force for
Stable Matching?
• What is the space complexity ??

st nd rd st nd rd
Man 1 2 3 Woman 1 2 3
X A B C A Y X Z
Y B A C B X Y Z
Z A B C C X Y Z

12
MEN PROPOSE (WOMEN DISPOSE)

Gale-Shapley Algorithm (men propose)

Initialize each person to be free.
while (some man m is free and hasn't proposed to every woman)
w = first woman on m's list to whom m has not yet proposed
if (w is free)

assign m and w as engaged

else if (w prefers m to her fiancé m')
assign m and w as engaged, and m' to be free
else
w rejects m

13
ANALYSIS
• Proof of Correctness
• Does the algorithm terminate?
• Matches everyone (Perfect matching) and
• Outputs a stable matching
• Running time and Space requirement Complexity?

Improvement Lemma: If a woman has a committed suitor, then she

will always have someone at least as good, from that point in time
onwards (and on the termination of the algorithm).
Corollary: Each woman will marry her absolute favorite of the men
who proposed to her.
Demotion Lemma: The sequence of women to whom a man m
proposes gets worse and worse (in terms of his preference list)
14
LEMMA 1
• No Man can be rejected by all the Women.
• Proof: by contradiction

Suppose Abebe is rejected by all the

women. At that point:
• Each women must have a suitor other than
Abebe (By Improvement Lemma, once a
woman has a suitor she will always have at
least one)
• The n women have n suitors, Abebe not among
them.
• Thus, there must be at least n+1 men !
15
COROLLARY OF LEMMA 1
• If a man m is free at some point in the execution of the algorithm,
then there is a woman to whom he has not yet proposed.

• The algorithm returns a perfect matching at the end. (Since there is

no free man or woman left.)

A perfect matching is a matching which covers

all vertices of the graph.

16
LEMMA 2
• Gale-Shapley algorithm returns a stable matching.
• Proof: by contradiction

Abebe Chala
Almaz
Martha
• Assume there is an unstable pair: Abeba and Martha
• This means Abebe likes Martha more than his partner, Almaz.
• Thus, Abebe proposed to Martha before he proposed to Almaz.
• Martha must have rejected Abebe for someone she preferred.
• By the Improvement lemma, she must like her partner Chala
more than Abebe.
• Pairs are not unstable. Contradiction!

17
QUESTION!
Who is better
off,
the men or the
women?

18
BEST (VALID) PARTNER FOR
Abebe?
• Best woman for “Abebe”? The woman at the top of Abebe’s list?
• A woman w is a valid partner of a man m if there is a stable
matching that contains (m,w).
• A man’s best valid partner is the highest ranked woman for
whom there is some stable pairing in which they are matched
• She is the best woman he can conceivably be matched in a stable
world.
• A Man’s worst valid partner is the lowest ranked woman in his
preference list that is a valid partner.

19
DATING DILEMMA
• A pairing is (woman-) man-optimal if every (woman) man gets
(her) his best valid partner. This is the best of all possible stable
worlds for every (woman) man simultaneously.

• A pairing is (woman-) man-pessimal if every (woman) man gets

(her) his worst valid partner. This is the worst of all possible
stable worlds for every (woman) man simultaneously.

• The Gale-Shapley algorithm always produces a man-optimal and

woman-pessimal pairing.

20
CONCLUSION: MARRY WELL!

21
ADVICE TO FEMALES

Learn to make the first move.

22
EXTENSIONS
Extensions
• Sets of unequal size
• Unacceptable partners
• Indifference
• Many-to-many assignments
Deceit, Coalitions and Strategy
• In general, lying cannot be totally discouraged in Stable Marriage
Problem. (sounds familiar??)
Other similar problems
• Internship assignments: Given a set of companies and students, pair
them.
• Given airlines and pilots, pair them.
• Computer Vision: Given two images, pair the points belonging to the
same point in 3D to extract depth from the two images.
• Dorm room assignments.
• Hospital residency assignments.

23
REVIEW PROBLEM
Gale-Shapley Algorithm (men propose)
Initialize each person to be free.
while (some man m is free and hasn't proposed to every woman)
w = first woman on m's list to whom m has not yet proposed
if (w is free)

assign m and w as engaged

else if (w prefers m to her fiancé m')
assign m and w as engaged, and m' to be free
else
w rejects m
Complexity and Space requirements?

24