Artificial Intelligence and Cyber Secutity Module 1.pptx

Fundamentals of Artificial
Intelligence & Cyber Security
Dr. Maryjo M George,
Asst. Prof.,
Dept. of AI & ML,
MITE, Moodabidri

MITE - Institute Vision & Mission

Knowledge (Remembering):
What is the capital of France?
Can you list the months of a year
Comprehension (Understanding):
Explain in your own words how photosynthesis works.
Summarize the main ideas of the book you just read.
Application (Applying):
Use the Pythagorean theorem to solve for the length of
the hypotenuse in a right triangle.

Analysis (Analyzing):
Compare and contrast the themes and characters in two
different novels.
Evaluation (Evaluating):
Critique the effectiveness of a government policy in
addressing poverty.
Synthesis (Creating):
Design a marketing campaign for a new product, including a
logo and advertising strategy.
Create a multimedia presentation on the cultural
significance of a traditional festival.

Continuous Internal Evaluation
CIE (50 marks)
Semester End Examination
SEE(50 marks)
3hrs, 100 marks
IA – 30 marks
1. IA1- 50 marks -1.5hrs
2. IA2- 50 marks -1.5hrs
Assignment- 20 marks
3. Assignment1 (10 marks)
4. Assignment2 (10 marks)
Min: 20/50 Min: 40/100

Copyright © 2019 by Wiley India Pvt. Ltd.,
4436/7, Ansari Road, Daryaganj, New Delhi-
110002

1.Voice Assistants: Voice-powered AI assistants like Siri (Apple), Google Assistant
(Android), and Bixby (Samsung) use natural language processing and machine
learning to understand and respond to voice commands and questions.
2.Camera Enhancements: AI is used for improving smartphone photography.
Features like scene recognition, facial recognition, and automatic adjustments to
settings like exposure and white balance are all AI-driven.
3.Predictive Text and Autocorrect: AI algorithms predict and suggest words or
phrases while you're typing, making texting and typing more efficient.
4.Voice Recognition for Texting: AI-powered speech-to-text technology enables
users to dictate messages, emails, and notes by speaking into their phones.
5.Smart Battery Management: AI algorithms monitor your usage patterns and
optimize battery life by closing background apps and adjusting power settings
accordingly.

Artificial Intelligence (AI) is a field with a broad scope, and its definition can
vary depending on who you ask. In simple terms, the main goal of AI is to
make computers perform tasks that typically require human intelligence.
Here are some reasons why people are interested in automating human
intelligence:
1.Understanding Human Intelligence: AI helps us better understand how
human intelligence works by trying to replicate it in machines.
2.Smarter Programs: AI aims to create smarter computer programs that
can learn, adapt, and make decisions like humans.
3.Solving Difficult Problems: AI techniques are valuable for solving
complex problems that are challenging for traditional computing methods.
Overview of Artificial Intelligence

To a layman:
AI may seem like the combination of "artificial" (something made by humans)
and "intelligence" (the ability to gain knowledge and use it).
For technical experts:
AI is a complex field that involves creating intelligent machines.
AI doesn't solely aim to mimic human intelligence; it's about building systems
that can solve problems, even if their methods are different from human
approaches
Intelligence vs. Artificial Intelligence:
•Intelligence: Natural, can increase with experience and is sometimes
hereditary. Doesn't rely on external electricity but on knowledge.
•Artificial Intelligence: Involves computer systems that need electrical energy
and a knowledge base to generate output. Knowledge in AI often comes from
human input.

•Strong AI: Claims that computers can reach human-level thinking and
reasoning. Aims for true machine intelligence that surpasses human
capabilities.
•Weak AI: Focuses on enhancing computers with intelligence for specific
tasks or domains. Such AI acts intelligently within its limited scope but
doesn't possess human-like general intelligence

Definitions of AI
Artificial Intelligence (AI) is a field where we make machines act smart and do things
that typically need human thinking. Let's break down the key ideas in simpler terms:
Four AI Properties:
•AI can be like humans in thinking and behaving.
•It can also think and act logically and rationally.
1. Acting Humanly (Turing
Test):
•Imagine chatting with a computer,
and you can't tell it's not a person.
That's smart AI.
•There's a contest (Loebner Prize)
to see if a computer can fool you
into thinking it's human.
Virtual Customer Support Chatbots

2. Thinking Humanly (Cognitive Modeling):
•It's not enough for AI to seem smart; it should think like humans.
•We study how humans think and try to make computers do the same using
experiments and computer tricks.
•There's a whole field (Cognitive Science) for this, mixing AI, psychology, linguistics,
and more
•ensemble human thinking and behavior, making them more effective and user-friendly
in various applications.

3. Thinking Rationally (Laws of Thought):
•AI can be like a super logical thinker, following strict
rules to make decisions.
•We use math and logic to teach computers to think
correctly.
•But, it's tough because real-life situations often need
more than just logic.
4. Acting Rationally (Rational Agents):
•AI can be like an intelligent agent with goals, like a
robot trying to achieve something.
•A smart agent acts in ways that maximize its goals,
but it doesn't have to know everything.
•It does its best with what it knows and the resources
it has.

Foundations of AI
Many older disciplines contribute to a foundation for artificial
intelligence.
Philosophy: logic, philosophy of mind, philosophy of science,
philosophy of mathematics
Mathematics: logic, probability theory, theory of computability
Psychology: behaviourism, cognitive psychology
Computer Science & Engineering: hardware, algorithms,
computational complexity theory
Linguistics: theory of grammar, syntax, semantics

Expert Systems
Expert systems are computer-based systems that mimic the decision-making
abilities of a human expert in a specific domain or field of knowledge.
Mass spectrograms are graphical representations of the masses of different ions
produced when a chemical compound is vaporized and ionized.

Early Expert Systems: Bridging Human Expertise
•Subsequent expert systems showcased AI's
potential:
• MYCIN (1975): Revolutionized medical
diagnosis, particularly in bacterial infections, by
mimicking the decision-making process of
medical experts.
• PROSPECTOR (1979): Unearthed molybdenum
deposits using geological data, highlighting the
capacity of AI in geology and resource
exploration.
• R1 (1982): Configured computers for Digital
Equipment Corporation (DEC), streamlining
complex technical tasks.

In 1981, Japan initiated a project known as the "Fifth Generation" project
with the goal of developing intelligent computers. This project was
significant because it aimed to advance the field of artificial intelligence
(AI) by focusing on the use of Prolog logic programming, a programming
language specifically designed for symbolic reasoning and knowledge
representation.
Here's what this development means:
1.The "Fifth Generation" Project: Japan's "Fifth Generation" project was
a government-funded research initiative that sought to achieve a new level
of AI capabilities. The term "Fifth Generation" referred to a vision of
computing technology that extended beyond traditional numerical
computation to include knowledge-based reasoning, natural language
understanding, and symbolic reasoning.
2.Prolog Logic Programming: Prolog is a programming language
particularly well-suited for tasks involving symbolic reasoning and
knowledge representation. It uses a formal logic called predicate logic to
represent and manipulate knowledge. Japan's decision to base its project
on Prolog indicated a commitment to symbolic AI and knowledge-based
systems.

1.MCC (Microelectronics and Computer Technology Corporation): In
response to Japan's ambitious "Fifth Generation" project, the United
States established the Microelectronics and Computer Technology
Corporation (MCC) in Austin, Texas, in 1984. MCC was a research
consortium comprising multiple American technology companies, and its
primary mission was to counter the Japanese project's advancements in
AI and semiconductor technology
certain limitations in the field of Artificial Intelligence (AI) that became
apparent during a period often referred to as the "AI Winter”.

certain limitations in the field of Artificial Intelligence (AI) that became apparent during a
period often referred to as the "AI Winter." Let's break down what these limitations
mean:
1.Brittleness:
1. In the context of AI, "brittleness" refers to the inability of AI systems to adapt to
unexpected or unusual situations.
2. AI systems, especially early ones like expert systems, were often designed to
perform well in specific, well-defined scenarios, but they struggled when faced
with inputs or conditions they hadn't been explicitly programmed or trained for.
3. This lack of adaptability made them "brittle" because they could break or
provide incorrect results when faced with novel or unforeseen circumstances.
2.Domain Specificity:
1. "Domain specificity" in AI refers to the limited scope or narrow focus of AI
systems.
2. Many early AI systems, including expert systems, were highly specialized and
could only excel within a predefined domain of knowledge.
3. They couldn't easily transfer their knowledge or skills to different domains or
tasks, which limited their versatility.
3.Knowledge Acquisition Bottleneck:
1. The "knowledge acquisition bottleneck" refers to the challenge of acquiring and
encoding expert knowledge into an AI system.
2. Building expert systems required substantial effort and resources to collect and
formalize the knowledge of human experts in a particular field.
3. This process was labor-intensive and time-consuming, often involving experts

The "AI Winter" is a historical period marked by reduced
funding and interest in AI research due, in part, to these
limitations. The field faced skepticism because early AI
systems, despite their successes, couldn't live up to the high
expectations of achieving human-level general intelligence.
However, it's essential to note that AI research eventually
emerged from these challenges and limitations, leading to
significant advancements in more recent years, with AI
systems that are less brittle, more adaptable, and capable of
tackling broader tasks and domains

• Artificial Intelligence is a broad field
– Different things to different people
– Multiple definitions are there by different people
• Main Objective of AI is
– Computers to do tasks that require human intelligence
• People want to automate human intelligence for following reasons
– Understand human intelligence better
– Smarter Programs
– Useful technique for solving difficult problem
• To a common man Artificial Intelligence is two words
– Artificial : Make a copy of something natural
– Intelligence : Ability to gain and apply knowledge and skills

• List of Task that requires Intelligence
– Speech generation & understanding
– Pattern Recognition
– Mathematical Theorem proving
– Reasoning
– Motion in obstacle filled space

Intelligence Vs Artificial Intelligence

Strong AI Vs Weak AI
• Weak AI – Helping people in their intelligence task
• Strong AI – Doing intelligence task on its own

Definition of AI
The Loebner Prize was an annual competition in
artificial intelligence that awarded prizes to the
computer programs considered by the judges to be the
most human-like

Acting Humanly – Turing Test
• The Turing test was developed by Alan Turing(A computer scientist) in 1950.
• He proposed that the “Turing test is used to determine whether or not a computer(machine)
can think intelligently like humans”?
• The Turing Test is a widely used measure of a machine’s ability to demonstrate human-like
intelligence.
The basic idea of the Turing
Test is simple:
•a human judge engages in a
text-based conversation with
both a human and a machine,
and then decides which of the
two they believe to be a
human.
•If the judge is unable to
distinguish between the
human and the machine
based on the conversation,
then the machine is said to
have passed the Turing Test.

Thinking Humanly : Cognitive Modeling
• Once we gather enough data, we can create a model to simulate the human process.
• This model can be used to create software that can think like humans.
• Of course this is easier said than done! All we care about is the output of the program given a particular
input.
• If the program behaves in a way that matches human behavior, then we can say that humans have a
similar thinking mechanism.
• Within computer science, there is a field of study called Cognitive Modeling that deals with simulating the
human thinking process.
• It tries to understand how humans solve problems.
• It takes the mental processes that go into this problem solving process and turns it into a software model.
• This model can then be used to simulate human behavior.
• Cognitive modeling is used in a variety of AI applications such as deep learning, expert systems, Natural
Language Processing, robotics, and so on.

Thinking Rationally : Laws of Thought

Acting Rationally : Rational Agents

Is automating intelligence is possible ?
• Answer lies in
– Turing Test
– Yes – its possible

Man Vs Computer
• AI is the study of how to make computers do
things which, at the moment, people do better
(Rich and Knight)
• Second definition
• Role of computer changes from something
useful to essential

Man Vs Computer
• What computers do better than people?
– Numerical computation
– Information Storage
– Repetitive operations
• What People can do better than computers?
– People have outperformed computers in activities which
involve intelligence.
– We do not just process information.
– We understand it , make sense out of what we see and hear
– Then we come out new ideas

List of characteristics intelligence should
possess

AI technique for mapping human knowledge

the relationship between cognitive science, computer modeling, and artificial
intelligence (AI). It also touches upon different perspectives on the goal of AI
and presents an alternative definition. Additionally, it mentions the importance
of symbolic and non-algorithmic methods in AI research.
Key points from the text include:
1.Continuous Process: Cognitive scientists develop theories of human
intelligence, which are then programmed into computer models by AI
researchers. These computer models are used to test and refine the validity of
these theories. This iterative process helps improve our understanding of
human intelligence.
2.Debate on AI Goals: There is a debate within the AI community about the
primary goal of AI. Some believe that AI's goal is to simulate intelligent behavior
on a computer, regardless of the techniques used. Others argue that true AI
should simulate intelligence using methods similar to those employed by
humans.
3.Alternative Definition: The text provides an alternative definition of AI as
"that branch of computer science which deals with symbolic, non-algorithmic
methods of problem solving." This definition highlights two key characteristics:
symbolic thinking, which is different from numerical processing, and non-
algorithmic problem-solving, which differs from the algorithmic approach

4.Focus on Symbolic and Non-Algorithmic Processing: AI
research places a significant emphasis on developing techniques that
involve symbolic reasoning and non-algorithmic problem-solving. This
approach aims to emulate human reasoning processes more closely.
5.Application Areas of AI: The text mentions that AI is applied in
various areas, including general problem-solving, expert systems,
natural language processing, computer vision, robotics, and other
domains. These applications demonstrate the wide-ranging impact of
AI in different fields.
In summary, the text highlights the ongoing relationship between
cognitive science and AI, the debate over AI goals, and the importance
of symbolic and non-algorithmic processing in AI research. It also
briefly mentions the diverse application areas of artificial intelligence.

Cognitive science to computer modeling

Application Areas of AI
• General Problem Solving
– Water jug, TSP, Tower of Hanoi etc.

Application Areas of AI
• Expert Systems

Natural Language Processing
NLP aims to enable
computers to
understand, interpret,
and generate human
language in a way
that is both valuable
and meaningful.

Example: Sentiment Analysis:
Determining and extracting the sentiment or emotional tone expressed in a piece of
text, such as a review, comment, or social media post
Review 1:* "I absolutely love this smartphone! The camera quality is amazing, and it's
so easy to use.“
In this review, the sentiment is clearly positive. NLP algorithms analyze the words and
context to understand that the customer is expressing satisfaction and delight about the
smartphone's features.
*Review 2:* "The phone constantly freezes, and the battery life is terrible. I regret
buying it.“
This review conveys a negative sentiment. NLP algorithms recognize the negative
language used ("freezes," "terrible battery life," "regret") and interpret the overall tone
as dissatisfaction.
*Review 3:* "The smartphone is functional. Nothing special, but it gets the job done.“
This review is neutral. The words used are neither strongly positive nor negative,
indicating a neutral sentiment.
In this example, NLP algorithms analyze the words, phrases, and context within these
reviews to automatically determine the sentiment expressed. Businesses often use
sentiment analysis to gauge customer opinions, feedback, and reviews at scale, enabling
them to make data-driven decisions based on customer sentiments.

• Natural Language Processing
Components of a Natural Language Computer: A computer capable of interacting
through natural language requires three key components:
•Parser: This component is responsible for breaking down sentences and
understanding their grammatical structure.
•Knowledge Representation System: This system stores information and knowledge
in a format that the computer can use to make sense of language input.
•Output Translator: It helps the computer generate meaningful responses in natural
language.

• Computer Vision
Computer vision is a field of artificial intelligence and computer science that focuses
on enabling computers to interpret and understand visual information from the world,
similar to how humans perceive and make sense of the visual world through their
eyes

Example: Face Recognition
Computer vision algorithms analyze the image to extract relevant features. In the
case of face recognition, these features might include the location of eyes, nose,
mouth, and other distinctive facial characteristics

AI system can learn to recognize evolving patterns of spam
and phishing attempts without relying solely on fixed rules.
Conventional spam filter operates based on rules like
checking for specific keywords, known spam email addresses,
AI computing vs conventional computing
Eg: Email Filtering

Intelligent Agent
Intelligent agents are computer programs or
systems designed to perform tasks
autonomously, adapt to changing
environments, and make decisions based on
their goals and available information

Intelligent agent
An agent is defined as something that sees (perceives) and acts in
an environment. IAs perform task that will be beneficial for the
business procedure, PC application or a person.

110002
Percept: It is defined as an agent’s perceptual inputs at a specified
instance.
Percept sequence: It is defined as the history of everything that an
agent has perceived till date.
Agent function
Agent Program

Vacuum agent
Perception:
1. Which room it is in
2. whether there is dirt in the room
Actions: move left, right, clean the dirt
Agent function: if the current room is dirty clean it
otherwise move to the next room

Agent function
Percept sequence Action
[A, Clean] Right
[A, Dirty] Suck
[B, Clean] Left
[B, Dirty] Suck
[A, Dirty], [A, Clean] Right
[A, Clean], [B, Dirty] Suck
[B, Dirty], [B, Clean] Left
[B, Clean], [A, Dirty] Suck
[A, Clean], [B, Clean] No-op
[B, Clean], [A, Clean] No-op

Agent function: It is defined as a map from the precept sequence
to an action.
Agent function, a = F(p)
where p : the current percept,
a : the action carried out,
F :the agent function
F maps precepts to actions
F :P → A where P :the set of all precepts
A : the set of all actions.
Generally, an action may be dependent of all the precepts
observed, not only the current percept, then,
F :P * → A

Agent program
.
Function REFLEX-VACUUM-AGENT([location, status]) returns Action
if status = Dirty then return Suck
if location= A then return Right
if location = B then return Left
return action

110002
REFLEX-VACUUM-AGENT is a function that takes an input consisting of two parameters:
location and status. location represents the current location of the vacuum agent (either A or
B), and status represents the status of the current location (either Clean or Dirty).
The code checks the status of the current location. If the status is "Dirty," the agent returns
the action "Suck" to clean the current location.
If the location is "A," the agent returns the action "Right" to move to location B.
If the location is "B," the agent returns the action "Left" to move to location A.
If none of the above conditions are met (i.e., if the location is clean and neither A nor B), the
agent returns a default action, which is "NoOp" (indicating no operation or staying in the
current location)

Structure of agent
The structure of agent can be represented as:
Agent= Architecture + Agent program
Program where,
Architecture : The machinery that an agent executes on
Agent program: An implementation of an agent function
Function REFLEX-VACUUM-AGENT([location, status])
returns Action
if status = Dirty then return Suck
if location= A then return Right
if location = B then return Left
return action

PAGE
While designing the intelligent systems, there are following four
main factors to be considered
• P, Percepts: The inputs to the AI system.
• A, Actions: The outputs of the AI system.
• G, Goals: What the agent is expected to achieve.
• E, Environment: What the agent is interacting with.
Copyright © 2019 by Wiley India Pvt. Ltd., 4436/7, Ansari Road, Daryaganj,
New Delhi-110002

Agent type Percepts Actions Goals
Enviro
nment
Medical
diagnostic
System
Symptoms,
test
results,
patient’s
answers
Questions,
test requests,
treatments,
referrals
Healthy
patients,
minimise the
costs
Patient,
hospital,
staff
Satellite
image
analysis
system
Pixels of
varying
intensity and
colour
Display a
categorisation
of
the scene
Correct image
categorisation
Images
from
orbiting
satellite
Part-picking
robot
Pixels of
varying
intensity and
colour
Pick up parts
and
sort them into
bins
Place parts
into
correct bins
Conveyo
r belt
with
parts,
bins
Refinery
controller
Temperature,
pressure and
chemical
readings
Open and close
valves, adjust
temperature
Maximise
purity,
yield, safety
Refinery
, staff
Interactive
English
Tutor Typed words
Display
exercises,
suggestions,
corrections
Maximise
exam results
Set of
students,
exam
papers

Part-
picking
robot
Pixels of
varying
intensity and
colour
Pick up parts
and
sort them into
bins
Place parts
into
correct bins
Convey
or belt
with
parts,
bins
Refinery
controller
Temperature,
pressure and
chemical
readings
Open and
close
valves, adjust
temperature
Maximise
purity,
yield, safety
Refiner
y, staff
Interactive
English
Tutor Typed words
Display
exercises,
suggestions,
corrections
Maximise
exam results
Set of
students
, exam
papers

Satellite image analysis system
Agent type Percepts Actions Goals Environment
Satellite image
analysis
system
Pixels of
varying
intensity and
colour
Display a
categorisation of
the scene
Correct image
categorisation
Images from
orbiting satellite

Part-picking robot
Part-picking
robot
Pixels of
varying
intensity and
colour
Pick up parts and
sort them into
bins
Place parts into
correct bins
Conveyor belt
with parts, bins

Refinery controller
Refinery
controller
Temperature,
pressure and
chemical
readings
Open and close
valves, adjust
temperature
Maximise
purity,
yield, safety
Refinery, staff
Interactive
English Tutor Typed words
Display exercises,
suggestions,
corrections
Maximise exam
results
Set of students,
exam papers

Agent program
On each invocation, the memory of the agent is updated to mirror
the new percept, the best action selected and the fact that the
action was taken is also stored inside the memory. The memory
persists from one invocation to the next.
Function SKELETON-AGENT(Percept) returns
Action
Static: Memory, the agent's memory of the world
Memory <- UPDATE-MEMORY(Memory, Percept)
Action <- CHOOSE-BEST-ACTION(Memory)
Memory <- UPDATE-MEMORY(Memory, Percept)
return Action

Memory <- UPDATE-MEMORY(Memory, Percept):
This line indicates that the agent updates its memory based on the percept it receives
from the environment. The function UPDATE-MEMORY is responsible for
incorporating new information from the environment into the agent's memory. This is
crucial because agents need to maintain an internal representation of the world to
make informed decisions.
Action <- CHOOSE-BEST-ACTION(Memory):
After updating its memory, the agent selects an action using the function CHOOSE-
BEST-ACTION. This function likely involves some form of decision-making process,
where the agent evaluates its current state and the information in its memory to
determine the most appropriate action to take. The "best" action might be based on
previously learned knowledge or a predefined strategy.
Memory <- UPDATE-MEMORY(Memory, Percept):
After choosing an action, the agent again updates its memory. This update may reflect
any changes in the environment due to the agent's chosen action.
return Action:
Finally, the agent returns the selected action as the output of the function, which it will
execute in the environment.

Attributes of agent
Autonomy: Agent work without the direct interference of the people or others
and have some kind of control over their action and the internal state.
Social ability: Agent’s interface with different agents and human by the means
or specific likeness of agent communication language.
Reactivity: An agent's ability to perceive and respond to its environment. Agents
perceive their condition which might include the physical world, client by the
means of graphical user interface, an accumulation of agent.
Proactivity: Proactivity is the opposite of mere reactivity. It means that an agent doesn't
just respond to stimuli from its environment but can also take initiative. Proactive agents
can set and pursue their own goals, make decisions independently, and exhibit goal-
directed behavior. This ability allows agents to plan, act, and achieve their objectives
even in the absence of external stimuli or directives.
Goal oriented: An agent is efficient of handling complex high-level tasks. The
decision for how such a task is best split into smaller subtasks, and in which
order and manner these subtasks should be composed by the agents itself.

Rationality
Rationality is defined as the only status of being sensible and with
great judgmental feeling. Rationality is only related to the expected
activities or actions and this results relying on what the IA has
viewed.
Performing activities with the point of obtaining the valuable data is
said to be a critical piece of rationality.
Rational Agent
A right action is always performed by a rational agent, where the right
action is equivalent to the action that causes the agent to be most
successful in the given percept sequence. The problem the agent
solves is characterised by the performance measure, environment,
actuators and sensors (PEAS).

Rational agent
Rationality of an agent depends on mainly on the following four
parameters:
• The degree of success that is determined by the performance
measures.
• The percept sequence of the agent that have been perceived till
date.
• Prior knowledge of the environment that is gained by an agent
till date.
• The actions that the agent might perform in the environment.

Ideal rational agent
An ideal rational agent is the one, which is competent enough
of performing expected actions to expand its performance
measure, on the basis of the following:
• Its percept sequence.
• Its built-in knowledge base.
Autonomy
If the behaviour of the system is determined by its own
experience, the system is known as autonomous. When the
agent has very less experience, it will be required to act in a
random manner unless some assistance is provided either by the
designer or the programmer. An actual autonomous IAs would
be able to successfully and efficiently adapt to all types of the
environments, given sufficient time.

Environment and its properties
1. Accessible (Fully observable) and Inaccessible Environments
(partially observable)
2. Deterministic Environment and Nondeterministic Environment
3. Episodic and Nonepisodic Environment
4. Static versus Dynamic Environment
5. Discrete versus Continuous Environment
6. Single Agent versus Multiagent Environment

Environment Types-I :
Fully observable (accessible) vs. partially
observable (inaccessible)
Environment Types-II: Deterministic vs.
stochastic (non-deterministic)
• Fully observable if agents sensors detect all
aspects of environment relevant to choice of
action
• Could be partially observable due to noisy,
inaccurate or missing sensors, or inability to
measure everything that is needed
• Model can keep track of what was sensed
previously, cannot be sensed now, but is probably
still true.
• Often, if other agents are involved, their
intentions are not observable, but their actions
are
• E.g chess – the board is fully observable, as are
opponent’s moves.
• Driving – what is around the next bend is not
observable (yet).
• · Deterministic = the next state of the
environment is completely predictable from the
current state and the action executed by the agent
• Stochastic = the next state has some uncertainty
associated with it
• Uncertainty could come from randomness, lack
of a good environment model, or lack of
complete sensor coverage
• Strategic environment if the environment is
deterministic except for the actions of other
agents
• Examples:
Non-deterministic environment: physical world:
Robot on Mars Deterministic environment: Tic
Tac Toe game

Environment Types-III : Episodic vs. sequential Environment Types-IV: Discrete vs. continuous
• The agent's experience is divided into
atomic "episodes" (each episode consists of the
agent perceiving and then performing a single
action) and the choice of action in each episode
depends only on the episode itself
• Sequential if current decisions affect future
decisions, or rely on previous ones
• Examples of episodic are expert advice systems –
an episode is a single question and answer
• Most environments (and agents) are sequential
• Many are both – a number of episodes containing
a number of sequential steps to a conclusion
• Examples:
Episodic environment: mail sorting system Non-
episodic environment: chess game
• Discrete = time moves in fixed steps, usually with
one measurement per step (and perhaps one action,
but could be no action). E.g. a game of chess
• Continuous = Signals constantly coming into
sensors, actions continually changing. E.g. driving a
car

Environment types-V:
Static vs. dynamic:
Environment types-VI
Single agent vs. multi agent:
• Dynamic if the environment may change over time.
Static if nothing (other than the agent) in the
environment changes
• Other agents in an environment make it dynamic
• The goal might also change over time
• Not dynamic if the agent moves from one part of
an environment to another, though it has a very
similar effect
• E.g. – Playing football, other players make it
dynamic, mowing a lawn is static (unless there is a
cat…), expert systems usually static (unless
knowledge changes)
• An agent operating by itself in an environment is
single agent!
• Multi agent is when other agents are present!
• A strict definition of an other agent is anything
that changes from step to step. A stronger definition
is that it must sense and act
• Competitive or co-operative Multi-agent
environments
• Human users are an example of another agent in a
system
• E.g. Other players in a football team (or opposing
team), wind and waves in a sailing agent, other cars
in a taxi driver

Fully
observable?
Deterministic? Episodic? Static? Discrete? Single
agent?
Solitaire No Yes Yes Yes Yes Yes
Backgammo
n
Yes No No Yes Yes No
Taxi driving No No No No No No
Internet
shopping
No No No No Yes No
Medical
diagnosis
No No No No No Yes
Crossword
puzzle
Yes Yes Yes Yes Yes Yes
English
No No No No Yes No

PEAS for Agent
automated Taxi Driver
Performance
Measure
Safety, time, legal
drive, comfort
Environment Roads, other cars,
pedestrians, road
signs
Actuators Steering, accelerator,
brake, signal, horn
Sensors Camera, sonar, GPS,
Speedometer,
odometer,
vacuum cleaner
Performance
Measure
cleanness, efficiency:
distance traveled to clean,
battery life, security
Environment room, table, wood floor,
carpet, different obstacles
Actuators wheels, different brushes,
vacuum extractor
Sensors camera, dirt detection
sensor, cliff sensor, bump
sensors, infrared wall
sensors

Types of agents
1. Simple reflex agent
2. Model based agent
3. Goal based agent
4. Utility based agent
5. Learning agent

1.Simple Reflex Agent:
1.A simple reflex agent is the most basic type of agent.
2.It makes decisions based solely on the current percept,
without considering past percepts or future consequences.
3.It uses a set of condition-action rules (if-then statements) to
map percept inputs to actions.
2.Model-Based Agent:
1.A model-based agent maintains an internal model of the
world or environment it operates in.
2.It uses this model to plan for future actions and make
decisions by considering the current state and its
understanding of how actions will affect that state.
3.Goal-Based Agent:
1.A goal-based agent has a specific goal or objective it is
trying to achieve.
2.It considers the current state, the desired goal state, and
plans a sequence of actions to reach the goal.

1.Model-Based Agent:
1.A model-based agent maintains an internal model of the
world or environment it operates in.
2.It uses this model to plan for future actions and make
decisions by considering the current state and its
understanding of how actions will affect that state.
2.Goal-Based Agent:
1.A goal-based agent has a specific goal or objective it is
trying to achieve.
2.It considers the current state, the desired goal state, and
plans a sequence of actions to reach the goal.
3.Goal-based agents are often more sophisticated and
capable of handling complex tasks.

5. Learning Agent:
1.A learning agent is capable of learning from its interactions
with the environment.
2.It can adapt and improve its performance over time
through learning algorithms, such as reinforcement
learning or supervised learning.
3.Learning agents can start with minimal knowledge and
gradually acquire knowledge and improve their decision-
making capabilities through experience.

4. Utility-Based Agent:
4. A utility-based agent is similar to a goal-based agent, but
it also assigns utilities or values to different states and
actions.
5. It aims to maximize the expected utility when making
decisions.
6. Utility represents the agent's preference for different
outcomes, and the agent tries to choose actions that
maximize expected utility.

motion-activated lights are simple reflex agents

2. Model-Based Agent:
Environment Model: A model-based agent maintains an internal model or
representation of the environment it interacts with. This model includes information
about the current state, possible future states, and the effects of different actions.
Comprehensive Knowledge: Model-based agents have a more comprehensive
understanding of the environment. They can anticipate the consequences of actions
by simulating them in their model.
Planning and Decision Making: These agents use their internal models to plan ahead.
They can consider multiple possible future states and choose actions that lead to
desirable outcomes. They may use algorithms like search and optimization to make
decisions.
Memory and History: Model-based agents typically have memory or a history of past
interactions, allowing them to consider the context of the current decision in light of
previous actions and observations.

2. Model based agent
Model-based agent is known as Reflex agents with an internal
state. One problem with the simple reflex agents is that their
activities are dependent of the recent data provided by their
sensors. On the off chance that a reflex agent could monitor its
past states (that is, keep up a portrayal or “model” of its history in
the world), and understand about the development of the world.

3. Goal based
Indeed, even with the expanded information of the current
situation of the world given by an agent’s internal state, the
agent may not, in any case, have enough data to reveal to it. The
proper action for the agent will regularly depend upon its
goals .Thus, it must be provided with some goal information.

4. Utility based agent
Goals individually are insufficient to produce top high-quality
behavior. Frequently, there are numerous groupings of actions
that can bring about a similar goal being accomplished. Given
proper criteria, it might be conceivable to pick ‘best’ sequence of
actions from a number that all result in the goal being achieved.
Any utility-based agent can be depicted as having a utility
capacity that maps a state, or grouping of states, on to a genuine
number that speaks to its utility or convenience or usefulness.

Types of agent
1. Simple reflex agent
Simple reflex agent is said to be the simplest kind of agent. These agents
select an action based on the current percept ignoring the rest of the
percept history.
These percept to action mapping which is known as condition-action
rules (so-called situation–action rules, productions, or if–then rules) in
the simple reflex agent. It can be represented as follows:
if {set of percepts} then {set of actions}
For example,
if it is raining then put up umbrella
or

Limitations
• Intelligence level in these agents is very limited.
• It works only in a fully observable environment.
• It does not hold any knowledge or information of
nonperceptual parts of state.
• Because of the static knowledge based; it’s usually too big to
generate and store.
• If any change in the environment happens, the collection of
the rules are required to be updated.

Updating status requires information on
1. How the world evolves independent of agent
2. What is impact of agent actions on enviornemnt

5. Learning agent
By actively exploring and experimenting with their environment,
the most powerful agents are able to learn. A learning agent can be
further divided into the four conceptual components

110002
Quiz 1
1.Explain the concept of intelligent agent
2.Explain any two types of environments in detail

Artificial Intelligence and Cyber Secutity Module 1.pptx

More Related Content

Similar to Artificial Intelligence and Cyber Secutity Module 1.pptx (20)

Recently uploaded (20)

Artificial Intelligence and Cyber Secutity Module 1.pptx

Editor's Notes