Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd
Upload
73 views

Algorithms PDF

Uploaded by

Mulika Fatima
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
73 views

Algorithms PDF

Uploaded by

Mulika Fatima
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 6

Algorithms

• Time to refocus on what we identified as the primary


object of study in computer science: the algorithm

• We never really stopped talking about algorithms; at


this point you can probably list a few:
Algorithms for interpreting bit sequences
The fetch-decode-execute algorithm
Algorithms for managing processes
Algorithms for network communication

Formal Definition
• We now move from our current intuitive understanding
of an algorithm to a formal one

• We start with a formal definition:


An algorithm is an ordered set of unambiguous,
executable steps that defines a terminating process

• The underlined words constitute the core ideas in this


definition, and separate the applied, informal uses of
the term from its use in stricter, more technical, and
scientific contexts
“What Something Is” vs.
“How It Is Written”
• The fundamental distinction of what an object is vs.
how that object is represented is crucial to computer
science, especially for algorithms

• Ultimately, an algorithm is abstract: it exists in its own


right (and in our minds) as a concept

• Thus, when we want to share an algorithm with


someone (or something) else, it is crucial that we
choose a representation that communicates the
algorithm completely and accurately — easier said than
done, as you might have noticed

Algorithms, Programs, and


Processes
• Some algorithm representations can only be
understood by humans: natural language (e.g., plain
English), sketches, diagrams

• A program is an algorithm representation that is meant


to be readable by a machine — which, today, typically
refers to the computer as we know it

• An instance of a computer’s performing or executing


that program is said to be a process — note that this is
precisely how the term is used in the context of
operating systems
Algorithm Representation
• We typically associate the act of representing concepts
with a language — some set of symbols (e.g., “A B C D
1 2 3 ; : . ! ?,” etc.) plus rules for putting them together
(e.g., “a sequence of letters without spaces makes a
word;” “a sequence of words ending with a period
makes a sentence;” “to represent the idea of a cat in
English, we put together the symbols c-a-t”)

• Algorithm representation is no different, though as an


object of formal study, some terms gain a precise
meaning than you might have seen previously

• The fundamental symbols that we use for representing


algorithms — analogous to our alphabet, numbers, and
punctuation, and thus forming the building blocks of
our representations — are called primitives

• These primitives are expected to be assembled in


accordance with certain rules, analogous to our natural
language grammar; this set of rules is called syntax

• When we put together these primitives and follow


these rules, our intent is to convey something
meaningful, such as the algorithm itself, or some of its
key steps and concepts — this is called semantics

• The total system may be called a programming language


Informally, the syntax of a programming language determines how a program in that
language looks, while the language’s semantics determines what that program means
Representation Examples
• Some algorithm representations are, for now,
understandable by humans only
Natural language
Diagrams (e.g., flowcharts), pictorial representations

• Pseudocode representation adds more formalism in


both notation and in meaning, but remains primarily
for human use

• When we finally represent the algorithm as a program


in a particular programming language, we have a form
that can be used by the computer

Key Elements of Any


Algorithm
Pseudocode notation crystallizes certain concepts that
are shared by almost all programming languages:
• Values and expressions — Notation for calculating and manipulating information, ranging from
simple values (numbers, letters, text) to more complex structures (lists, data with distinct
properties [e.g., an address has a number, street, city, and zip], and arbitrary combinations of
these items [e.g., an address book of contacts, each with a list of addresses])

• Naming and storage — Notation for “holding on” to certain values and referring to them in
various parts of an algorithm (e.g., variables like x, index, or valueList)

• Sequencing — Clear indication of the order in which an algorithm’s steps will occur
• Conditions and branching — Different paths that an algorithm may follow based on some
decision or state

• Loops — Repetition of a particular activity or set of activities, either depending on some


condition or for a predetermined number of times (e.g., entering a password up to three
times only; performing something once for every item in a list)
Algorithm Discovery
• All of this talk about representing an algorithm assumes
that we even have an algorithm — and it’s true that for
many endeavors, algorithms are known
Converting temperature

Calculating an average or mean

Making change

Playing rock-scissors-paper…and many more

• Note how the struggle so far has been all about


representation — in particular, how do we express
these algorithms so that a machine can perform them

…but that’s just a fraction of the fun!


• Representing an algorithm for a machine is called
programming, and this is the activity that most people
associate with computer science
• But think about it — if we don’t know what the
algorithm is in the first place, then there wouldn’t be
anything to program at all!
• In the end, the core innovations in computer science
come from the formulation of an algorithm —
independently of any programming language
How to manage multiple processes in an operating system
How to enable billions of computers to communicate
How to identify spam

• To paraphrase the Bard: “The algorithm’s the thing”


Approaches to Algorithm
Discovery
• In the end, there is no single guaranteed recipe for
discovering an algorithm (think about that for a
moment…that is like saying there is no known
algorithm for discovering another algorithm!)
• But there are some general strategies that tend to
increase our chances of finding such an algorithm
• Polya’s approach specifies four phases: understand the
problem, form plan(s) to solve the problem, execute
the plan(s), then evaluate their effectiveness
Einstein was once asked what he would do if he was given five minutes to solve a
problem; he said he would spend the first four minutes just reading the problem

• Two complementary philosophies have emerged for


tackling computing problems:
Top-down — Break a problem up into smaller subproblems, then solve those first

Bottom-up — Look at specific instances of a problem, then see if those individualized


solutions can lead you to a general answer

Successful algorithm discovery usually takes a little bit


of both approaches
• Analogous approaches can be found in techniques of
mathematical proof:
Induction — Start with the simplest possible case, then build up a solution incrementally

Contradiction — See if you can solve an “opposite” problem, then “invert” the solution

• In the end, algorithm discovery remains very much an


art, as evidenced by the wealth of algorithms that are
yet to be discovered

You might also like