International Higher Diploma in Computer Studies
Artificial Intelligence
The marks given in brackets are indicative of the weight given to each part of the question.
Answer FOUR questions out of SIX.
Time: TWO hours and 10 minutes reading time
Reference materials are NOT allowed.
Question 1
a)
Define Artificial Intelligence.
[5 Marks]
Artificial Intelligence (AI) is a branch of computer science concerned with the design and
implementation of programs, which can simulate the human skills such as problem solving,
visual perception and language understanding.
b)
With the help of diagrams, differentiate between Breadth-First Search and
Depth-First Search techniques.
[10 Marks]
Breadth-First Search (BFS)
Breadth-first search goes through the tree level by level, visiting all of the nodes on
the top level first, then all the nodes on the second level, and so on. This strategy has
the benefit of being complete (if there's a solution, it will be found), and optimal as
long as the shallowest solution is the best solution.
Depth-First Search (DFS)
Depth-first search goes through the tree branch by branch, going all the way down to the leaf
nodes at the bottom of the tree before trying the next branch over. This strategy requires
much less memory than breadth-first search, since it only needs to store a single path from
the root of the tree down to the leaf node.
IHDCS/AI/0112
Page 1 of 8
c)
Differentiate between object recognition and pattern recognition.
[10 Marks]
Object Recognition stage is the ultimate goal of many vision systems. Its main uses include:




move around safely avoiding objects,
pick and place various objects,
inspect objects
perform many other tasks.
Pattern Recognition
“Pattern recognition is the research area that studies the operation and design of systems that
recognize patterns in data. It encloses subdisciplines like discriminant analysis, feature
extraction, error estimation, cluster analysis (together sometimes called statistical pattern
recognition), grammatical inference and parsing (sometimes called syntactical pattern
recognition). Important application areas are image analysis, character recognition, speech
analysis, man and machine diagnostics, person identification and industrial inspection.”
Question 2
a)
What is the objective of alpha beta pruning technique?
[8 Marks]
It is a technique that improves upon the minimax algorithm by ignoring branches on the game
tree that do not contribute further to the outcome. The basic idea behind this modification to
the minimax search algorithm is the following. During the process of searching for the next
move, not every move (i.e. every node in the search tree) needs to considered in order to
reach a correct decision. In other words, if the move being considered results in a worse
outcome than our current best possible choice, then the first move that the opposition could
make which is less then our best move will be the last move that we need to look at.
A subtree is pruned when the algorithm finds an alpha or beta that is not as good as what has
been found so far.
IHDCS/AI/0112
Page 2 of 8
If you assume that the algorithm is set to evaluate the best nodes first, it changes the effective
branching factor from b, to the square root of b, allowing the algorithm to go about twice as far as
minimax in the same time.
b)
Discuss the problem arising from machine vision.
[8 Marks]
People can easily make sense of what they see around them, easily recognising complex
objects - it's something we learn when we are very young. However, like natural language
understanding, this is extremely hard to automate. It requires both knowledge of objects in
the world (e.g., cats are furry and have a tail), knowledge of certain basic properties of the
physical world (e.g., objects generally have continuous smooth surfaces), and knowledge of
basic optics (e.g., image intensity depends on the reflectance of the object). Recognition of
objects is complicated by the fact that a single object may be viewed in many different ways,
light and shadows may be different, other objects may be in front of it, and so on.
c)
What do you understand by blind search technique?
[5 Marks]
The method is the blind search - a search for the best answer at random. Pure blind search is
usually simple to run, and therefore fast to realize. It often finds answers that are good
enough for practical purposes, or at least can serve as the preliminary estimates. Various ad
hoc modifications increase accuracy and are usually easy to implement, too.
Blind search, also called uninformed search, works with no information about the search space, other
than to distinguish the goal-state from all the others. These algorithms do not have any knowledge as
to which “direction” the Goal State lies.
d)
List down four areas under AI.
[4 Marks]
Natural Language
Problem Solving
Pattern Recognition
Automatic Programming
Game Playing
Expert System
Neural Networks
Robotics
Question 3
a)
List and describe different types of robots.
[12 Marks]
Tabletop Robots
Tabletop robots are robots that are small enough to fit onto a ping-pong sized table even
though many of them are designed to run on the floor in spite of their size.
Mid-size Robots
Mid-sized robots are generally a bit bigger and although they can be set on a table or
workbench for construction/maintenance they are too big to be tested there and must live on
the floor.
IHDCS/AI/0112
Page 3 of 8
Members Robots:
'To big for the Kitchen' robots are generally just that. They are designed for industrial settings
and are much more complex and expensive. They often require vision systems, and people
protection circuits as they can hurt someone if they go astray.
Air and Sea Robots
Air and Sea robots are generally designed to fly or float. There are whole sets of unique
problems for each of these.
b)
Describe five methods which are used in machine learning.
[8 Marks]
Inductive Learning
This refers to learning which progresses from the specific to the general. Used in knowledge
acquisition.
Case-based reasoning and analogical reasoning
This is defined as knowledge and/or inferences, which are derived from case history and
analogies. It is used in knowledge acquistion and in inferencing.
Statistical methods
It is used in knowledge acquisition and problem solving.
Genetic algorithms
The goal of genetic algorithms is to develop systems that can demonstrate self-organization,
adaptation and learning by experience through exposure to the environment. This is similar
but simplied to the way in which biological systems learn.
Neural Nets
A model that imitates biological neural network structures. The software simulates massively
parallel processes which involve processing elements called artificial neurons interconnected
in a network structure.
c)
What is Natural Language Processing?
[5 Marks]
The term: “natural” languages refer to the languages that people speak, like English and Japanese and
Swahili, as opposed to artificial languages like programming languages or logic. Natural Languages
are languages used in human culture such as Chinese, English or Bulgarian. They can be either
spoken or written.
NLP (Natural Language Processing) is a field in AI involving anything that processes natural
language. Extensive research in NLP over the past decade has brought us one of the most useful
applications of AI: machine translation. NLP research also deals with speech recognition. Currently,
programs that convert spoken speech into text have been widely used and are fairly dependable.
IHDCS/AI/0112
Page 4 of 8
Question 4
Write short notes on the following: a) Pattern Recognition
“Pattern recognition is the research area that studies the operation and design of systems that
recognize patterns in data. It encloses subdisciplines like discriminant analysis, feature
extraction, error estimation, cluster analysis (together sometimes called statistical pattern
recognition), grammatical inference and parsing (sometimes called syntactical pattern
recognition). Important application areas are image analysis, character recognition, speech
analysis, man and machine diagnostics, person identification and industrial inspection.”
b) Game Playing
Game playing is a search problem where you can have perfect decisions and imperfect decisions.
Alpha-Beta Pruning is a technique to eliminate parts of the game tree. Games are easy to represent on
a computer. Game-playing agents typically have a small number of actions. It can be considered an
idealization of the real world since you have to interact with others.
c) Robotics
Robot is a programmable machine that imitates the actions or appearance of an intelligent creature–
usually a human. To qualify as a robot, a machine has to be able to do two things: 1) get information
from its surroundings, and 2) do something physical–such as move or manipulate objects.
Robotics and AI/ALife are often seen as two different fields entirely, robotics being a mechanical
engineering field, and AI/ALife being computer science related. Whilst this is very true, robotics and
AI are closely meshed, both in obvious, and less obvious ways.
d) Expert Systems
Expert Systems are computer programs that are derived from a branch of computer science
research called Artificial Intelligence (AI). AI's scientific goal is to understand intelligence by
building computer programs that exhibit intelligent behavior. It is concerned with the
concepts and methods of symbolic inference, or reasoning, by a computer, and how the
knowledge used to make those inferences will be represented inside the machine.
e) Knowledge Acquisition
[5x5 Marks]
Knowledge acquisition refers to the task of endowing expert systems with knowledge, a task
currently performed by knowledge engineers. The choice of reasoning method, or a shell, is
important, but it isn't as important as the accumulation of high-quality knowledge. The power
of an expert system lies in its store of knowledge about the task domain -- the more
knowledge a system is given, the more competent it becomes.
IHDCS/AI/0112
Page 5 of 8
Question 5
a)
What is Genetic Programming?
[5 Marks]
Genetic programming is an excellent way of evolving algorithms that will map data to a
given result when no set formula is known. Mathematicians/programmers could normally
find algorithms to deal with a problem with 5 or so variables, but when the problem increases
to 10, 20, 50 variables the problem becomes close to impossible to solve.
b)
Define “Frames” as a knowledge representation method.
[5 Marks]
Frames are a variant of nets that are one of the most popular ways of representing nonprocedural knowledge in an expert system. In a frame, all the information relevant to a
particular concept is stored in a single complex entity, called a frame. Superficially, frames
look pretty much like record data structures. However frames, at the very least, support
inheritance. They are often used to capture knowledge about typical objects or events, such as
a typical bird, or a typical restaurant meal.
c)
Briefly explain the concepts of Inductive Learning.
[5 Marks]
This refers to learning which progresses from the specific to the general. Used in knowledge
acquisition.
Inductive learning systems learn by analyzing examples to identify correlations between inputs and
outputs. For example, neural network models process inputs according to networks of idealized
neurons, and learn by algorithms that adjust the weights of neural connections based on correlations
between inputs and outputs in training examples.
d)
Differentiate between Neural Networks and Convention computers.
[10 Marks]
Neural Networks are essentially a type of computer, but it does not work in the same way as
the conventional computer.
Neural Network it has very different computational properties. The conventional computer
has two main components, a memory and some kind of processing device (that is a CPU).
Information is represented in terms of structures of symbols, and the way in which this
information is processed depends on a program also stored in the computer's memory.
The neural network on the other hand is made up of a set of simple processing units connected
together in a network. The connections between the units can be thought of as wires, which carry
electrical activation from one unit to another. The units themselves have some similarity to biological
neurons. They store and pass on a certain amount of electrical activation that is related to the amount
of activation they themselves receive through their input connections. Information is represented with
levels of activation and the way in which it is processed all depends on the way in which activation
propagates through the network; i.e. it all depends on the connections between the units (hence the
name connectionism).
IHDCS/AI/0112
Page 6 of 8
Question 6
a) Explain the following Rule-Based systems.
i) Forward Chaining Systems
Forward-chaining means that the system begins with the axioms and rules, then reviews conclusionsmuch like one might prove a theorem in geometry.
Forward chaining is reasoning from known knowledge (or data) to new knowledge. It might be used
to match the customer's behavior to patterns describing appropriate lifestyles and banking products. If
our banking assistant were implemented using forward-chaining, it might contain rules such as
“parents with high incomes and young children often want college savings accounts.” Forwardchaining is ideal for applications with a large amount of data, such as sensor processing. Forwardchaining will provide all the conclusions a system can possibly reach unless specifically halted.
ii) Backward Chaining Systems
A backward-chaining system begins with a hypothesis to be proved, and then proceeds to determine
what the system must know in order to prove it.
Backward-chaining reasoning tends to be more focused than forward-chaining. Backward-chaining is
also more appropriate for diagnostic applications, such as the banking assistant. Such applications
tend to have high-level goals and causes of the goals (e.g., a desire to minimize fees), which must be
considered.
b) Define “Syntax” and “Semantics”.
[2 x 5 Marks]
[5 Marks]
Syntax
These words group themselves together into phrases, in these phrases in turn combine into sentences.
This is the level of syntax.
Syntax helps us understand how words are grouped together to make complex sentences, and gives us
a starting point for working out the meaning of the whole sentence.
Semantics
The problem of how to represent the meaning of sentences is undertaken in the level of semantics.
In general, the input to the semantic stage of analysis may be viewed as being a set of
possible parses of the sentence, and information about the possible word meanings. The aim
is to combine the word meanings, given knowledge of the sentence structure, to obtain an
initial representation of the meaning of the whole sentence.
IHDCS/AI/0112
Page 7 of 8
c) With the help of illustration, discuss how Expert System is designed.
[10 Marks]
The user interacts with the system through a user interface, which may use menus, natural language
or any other style of interaction). Then an inference engine is used to reason with both the expert
knowledge (extracted from our friendly expert) and data specific to the particular problem being
solved. The expert knowledge will typically be in the form of a set of IF-THEN rules. The case
specific data includes both data provided by the user and partial conclusions (along with certainty
measures) based on this data. In a simple forward chaining rule-based system the case specific data
will be the elements in working memory.
Almost all expert systems also have an explanation subsystem, which allows the program to
explain its reasoning to the user. Some systems also have a knowledge base editor, which
help the expert or knowledge engineer to easily update and check the knowledge base.
IHDCS/AI/0112
Page 8 of 8