COMP90038 Algorithms and Complexity

Introduction and Welcome

Junhao Gan
(with thanks to Harald Søndergaard, Toby Murray and Olya
Ohrimenko)

Week 1 Lecture 1

Semester 1, 2023

1/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 1 / 27
 Welcome to COMP90038

In this subject, we will talk about:

Data structures, including arrays, linked lists, stacks, queues, trees, priority
queues and graphs.

Algorithms for various problems, including sorting, searching, string manipu-

lation, graph manipulation, and more.

Algorithmic techniques, including brute force, decrease-and-conquer, divide-

and-conquer, dynamic programming and greedy approaches.

Analytical and empirical assessment of algorithms.

Complexity classes.

2/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 2 / 27
 Textbook

Anany Levitin. Introduction to the Design and Analysis of Algorithms. 3rd

Edition, Pearson, 2012.

Also available on the LMS, under “Reading Resources”:

Steven Skiena. The Algorithm Design Manual. Springer, 2008.

3/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 3 / 27
 Subject Coordinator

Lecturer and subject coordinator: Junhao Gan

Research Interests:
Practical algorithms with non-trivial theoretical guarantees for solving problems on
massive data.
“practical” means: the algorithms must be light-weighted, relatively easy-to-
implement and run fast in practice.
“non-trivial theoretical guarantees” means the algorithms must either beat or
match the theoretical bounds of the state-of-the-art solutions.
“massive data” means the datasets we deal with can be massively large, some-
times they may not be stored in the main memory or even not in a single
machine.

I am also a co-lecturer (with Prof. Tony Wirth) of COMP90077 Advanced

Algorithms and Data Structures in this semester, an advanced version of this
subject.
4/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 4 / 27
 Tutors and Canvas

Head tutor: Lianglu Pan

Tutors:
Philip Cervenjak
Zifeng (Peaker) Guo
Zhuowei (Alvin) Zhao
Yuning Zhou
possibly one more tutor

Other support is provided by your classmates, for example via the Discussion Board
in Canvas (LMS).
The Canvas page is our notice board, repository, and discussion forum.

5/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 5 / 27
 Lectures and Tutorials

We have two consecutive 1-hour lectures (i.e., a 2-hour lecture) in the Sunderland
Theatre on Wednesdays.

We use lecture capture which is useful for revisiting points from a lecture; and it
can, sort of, be a substitute for the lecture proper, even if it is a poor one.

Tutorials start in Week 2.

6/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 6 / 27
 Assessment

10 Weekly Quizzes: each worth 1% and in total worth 10%

Assignment 1: due around Week 6, worth 15%
Assignment 2: due around Week 11, worth 15%
A 3-hour final exam: worth 60%

Hurdles: To pass the subject you must obtain:

at least 30/60 in the written exam and

at least 50/100 overall.

All these details, and more, can be found on Canvas, under “Subject Overview”.

7/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 7 / 27
 Time Commitment

For the 12 weeks of semester, expect to spend

34 hours in lectures + tutorials,

30 hours on assignments
24 hours of reading and reviewing
24 hours of tutorial preparation
8 hours on quizzes and discussion

On average: 10 hours per week

The commitment is well worth it:

Knowledge of algorithms is essential for any computing professional.

It expands your mind, improves complexion, and contains all the minerals and
vitamins essential for developing boundless wisdom.
8/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 8 / 27
 Assumed Knowledge

There are two diagnostic quizzes for Week 1

not compulsory, but to help you work out how well you know the assumed
background knowledge
Mathematics: sets, relations, functions, recurrence relations
Programming: arrays, records, linked lists, dictionaries, functions,
procedures, formal and actual parameters, parameter passing, return-values,
pointers/references

See the Reading Resources page on Canvas which has some pointers to online
resources that may help with the above background knowledge

9/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 9 / 27
 How to Succeed

Understand the material (by doing), DO NOT just memorize it

If you fall behind, try to catch up as fast as possible.

Don’t procrastinate. Start assignments early. Put in the necessary time.

Attempt the tutorial questions every week, before you attend the tutorial, if at all
possible.

10/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 10 / 27
 How to Succeed

Support the learning of your fellow students and expect their support, in class and
through the LMS discussion board.

We are all on the same learning journey and have the same goal.

Participate in the discussions on the subject’s LMS site and check regularly for
announcements.

11/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 11 / 27
 Over to You—Introductions

Please introduce yourself to your neighbours.

Tell them where you are from, what degree program you are enrolled in, which
languages or programming languages you speak, what music you listen to, whatever.

12/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 12 / 27
 Over to You—A Maze Problem

A maze (or labyrinth) is contained in a 10 × 10 rectangle; rows and columns are

numbered from 1 to 10.

It can be traversed along rows and columns: up, down, left, right.

The starting point is (1,1), the goal point is (10,10).

These points are obstacles that you cannot travel through:

(3,2) (6,6) (2,8) (5,9) (8,4) (2,4) (6,3) (9,3) (1,9) (3,7)
(4,2) (7,8) (2,2) (4,5) (5,6) (10,5) (6,2) (6,10) (7,5) (7,9)
(8,1) (5,7) (4,4) (8,7) (9,2) (10,9) (2,6)

Find a path through the maze.

13/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 13 / 27
 14/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 14 / 27
 What is a Problem?

The Oxford Dictionary says: “doubtful or difficult question or task.”

In computer science we use the term like that too, but there is a more technical
concept of algorithmic problem.

We usually want to find a single generic solution to a bunch of similar questions.

For example, the “maze problem” is to come up with a mechanical solution to

any particular maze.

So to us, a “problem” usually has many (input) instances, sometimes infinitely

many.

15/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 15 / 27
 Example Definitions of the Maze Problem

Consider a maze contained in size 10×10 rectangle; rows and columns are numbered
from 1 to 10. One can move along rows and columns: up, down, left, right.

Given a list of specified obstacle coordinates, a start point coordinate and a goal
point coordinate, decide if there exists a path from the start point to the goal point.

Question 1: Are there infinitely many input instances? And why?

Question 2: How about we further parameterize the size of the rectangle as

α × α, where α is a positive integer?

16/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 16 / 27
 Algorithmic Problems

So a problem in computer science typically means a family of instances of a general

problem.

An algorithm for the problem has to work for all possible (input) instances.

Example: The sorting problem—an instance is a sequence of items.

Example: The graph colouring problem—an instance is a graph.

Example: Equation solving problems—an instance is a set of, say, linear equations.

17/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 17 / 27
 What is an Algorithm?
The dictionary: “process or rules for (esp. machine) calculation etc.”

A finite sequence of instructions

No ambiguity, and each step is precisely defined

Should work for all (well-formed) input
Should finish in a finite (reasonable) amount of time

The (single) description of a process that will transform arbitrary input to the
correct output—even when there are infinitely many possible inputs.

Like a cookbook recipe?

Sort of, but more like a method, a systematic approach that works for any instance.

For example, the Depth-First Search (DFS) algorithm can solve the Maze Problem
(with infinitely many input instances).
18/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 18 / 27
 What is an Algorithm?

Not long ago, “algorithm” was synonymous with “numeric algorithm”.

Mathematicians had found many clever algorithms for all sorts of numeric problems.

In BC 300, the ancient Greek mathematician Euclid proposed the first algorithm for
calculating the greatest common divisor (GCD) of two given positive integers
m and n.

The Euclidean Algorithm gcd(m, n) works as follows:

Step 1: If n = 0, return the value of m as the answer and stop.

Step 2: r ← m mod n.
Step 3: m ← n and n ← r;
go to Step 1.

Nowadays, numeric algorithms are just a small part of the syllabus in an algorithms
course.

19/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 19 / 27
 Computability

This year, 2023, is Alan Turing’s 111th birthday.

At the time of Turing’s birth, a “computer” was a hu-
man employed to do tedious numerical calculations.
Legacy: “Turing machine”, the “Church-Turing the-
sis”, “Turing reduction”, the “Turing test”, the “Turing
award”

One of Turing’s great accomplishments was to put the concept of an algorithm

on a firm foundation and to establish that certain important problems do not
have algorithmic solutions, e.g., the Halting Problem.

20/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 20 / 27
 Abstract Complexity

In this subject, we are only interested in problems that have algorithmic solutions.

However, amongst those, there are many that provably do not have efficient solu-
tions.

Towards the end of this subject, we will discuss complexity theory briefly—this
theory is concerned with the inherent “hardness” of problems.

21/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 21 / 27
 Why Study Algorithms?

Computer science is increasingly an enabler for other disciplines, providing useful

tools for these.

Algorithmic thinking is relevant in life sciences, in engineering, in linguistics, in

chemistry, etc.

Today computers allow us to solve problems whose size and complexity is vastly
greater than what could be done a century ago.

The use of computers has changed the focus of algorithmic study completely,
because algorithms that work well for a human (small scale) usually do not work
well for a computer (big scale).

22/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 22 / 27
 Why Study Algorithms and Their Complexity?

To collect a number of useful problem solving tools.

To learn, from examples, strategies for solving computational problems.

To be able to write robust programs whose behaviour we can reason about.

To develop analytical skills.

To learn about the inherent difficulty of some types of problems.

23/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 23 / 27
 Problem Solving Steps

Understand the problem

Decide on the computational means (sequential/parallel, exact/approximate)
Decide on method to use (algorithm design technique or strategy, use of
randomization)
Design the necessary data structures and algorithm
Check for correctness, trace example input
Evaluate analytically (time, space, worst case, average case)
Code it
Evaluate empirically

24/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 24 / 27
 What We Will Study

Algorithm analysis
Important algorithms for various problems, primarily
Sorting
Searching
String processing
Graph algorithms

Approaches to algorithm design

Brute force
Decrease and conquer
Divide and conquer
Transform and conquer

25/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 25 / 27
 Study Tips

Before the lecture, as a minimum make sure you have read the introductory section
of the relevant chapter.

Always read (and work) with paper and pencil ready; run algorithms by hand.

Always have a go at the tutorial exercises; this subject is very much about learning-
by-doing.

After the lecture, re-read and consolidate your notes.

Identify areas not understood and use the LMS Discussion Forum.

Rewrite your notes if that helps.

26/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 26 / 27
 Things to Do in the First Two Weeks

Get the textbook, read Chapter 1, and skim Chapter 2.

Make sure you have a unimelb account.

Visit the COMP90038 LMS pages and check the weekly schedule and any new
announcements.

Use the LMS Discussion Board; for example, if you are interested in forming a
study group with like-minded people, the Discussion Board is a useful place to say
so.

27/27
Algorithms and Complexity (Sem 1, 2023) Introduction and Welcome © University of Melbourne 27 / 27