SUB: ARTIFICIAL INTELLIGENCE (KCS071)

VII-SEM
FACULTY NAME:SONALI SINGH

UNIT-1(INTRODUCTION)

S.NO TOPIC NAME

1 Introduction
2 Definition
3 Future of Artificial Intelligence
4 Characteristics of Intelligent Agents
5 Typical Intelligent Agents
6 Problem Solving Approach to Typical AI
problems.
1.INTRODUCTION
Artificial intelligence (AI) is the theory and development of computer systems capable of
performing tasks that historically required human intelligence, such as recognizing speech,
making decisions, and identifying patterns. AI is an umbrella term that encompasses a wide
variety of technologies, including machine learning, deep learning, and natural language
processing (NLP).
Artificial Intelligence (AI) refers to the simulation of human intelligence processes by machines,
particularly computer systems. These processes include learning (acquiring information and rules
for using it), reasoning (using rules to reach approximate or definite conclusions), and
self-correction.
The four types of artificial intelligence
categorized by their capabilities are:
1. Reactive Machines

● Definition: These AIs do not have memory or the ability to learn from past
experiences. They operate solely based on current inputs.
● Example: IBM’s Deep Blue, which played chess by evaluating possible
moves in real-time without learning from past games.

2. Limited Memory

● Definition: These systems can learn from historical data and use that
information to make decisions. They have a form of memory but are still
task-specific.
● Example: Self-driving cars, which collect and analyze data from previous
trips to improve navigation and safety.

3. Theory of Mind

● Definition: This type of AI aims to understand human emotions, beliefs, and

social dynamics. It is still largely theoretical and in development.
● Example: AI systems that can interpret emotional cues and engage in more
human-like interactions.

4. Self-aware AI

● Definition: The most advanced form of AI, where systems possess

self-awareness and consciousness. This type is still hypothetical and poses
significant ethical questions.
● Example: Currently, there are no real-world examples, as this level of AI
has not yet been achieved.
2.Definition
several definitions of Artificial Intelligence (AI) from various prominent scientists and
researchers in the field:

1. John McCarthy (1956):

○ Often referred to as the "father of AI," McCarthy defined AI as "the science
and engineering of making intelligent machines."
2. Alan Turing (1950):
○ Turing described AI in terms of the ability of machines to exhibit behavior
indistinguishable from that of a human, famously proposing the "Turing
Test" as a measure of machine intelligence.
3. Marvin Minsky (1967):
○ Minsky defined AI as "the science of making machines do things that
would require intelligence if done by men," emphasizing the replication of
human cognitive functions.
4. Herbert Simon (1957):
○ Simon defined AI as "the study of the processes of human thinking,"
suggesting that AI systems should mimic human thought processes.
5. Ray Kurzweil (1999):
○ Kurzweil defines AI as "the ability of a computer or other machine to
perform tasks that are usually associated with intelligent beings,"
highlighting the practical applications of AI.
6. Stuart Russell and Peter Norvig (2009):
○ In their book "Artificial Intelligence: A Modern Approach," they define AI as
"the study of agents that receive precepts from the environment and take
actions that maximize their chances of success."
7. Geoffrey Hinton:
○ Known as a pioneer of deep learning, Hinton describes AI as "the ability of
a machine to perform tasks that, when done by humans, require
intelligence."
8. Judea Pearl:
○ Pearl focuses on the reasoning aspect, stating that AI is "the study of the
computation that makes it possible to make inferences."

These definitions highlight the diversity in understanding AI, encompassing its

theoretical foundations, practical applications, and goals of mimicking or replicating
human intelligence.
overall Artificial Intelligence (AI) refers to the field of computer science that focuses on
creating systems capable of performing tasks that typically require human intelligence.
These tasks include reasoning, learning, problem-solving, understanding natural
language, and perception.

Historical Background of Artificial

Intelligence
The evolution of Artificial Intelligence (AI) is marked by significant milestones and
contributions from various researchers and institutions. Here’s a chronological overview
of key developments in the history of AI:

1. Early Concepts (Pre-1950s)

● Philosophical Foundations: Ideas about automata and machines that mimic
human thought can be traced back to ancient philosophers, such as Aristotle,
who explored logic and reasoning.
● Computational Theories: In the 19th century, figures like George Boole and
Gottfried Wilhelm Leibniz laid groundwork for formal logic and binary systems.

2. Birth of AI (1950s)
● Alan Turing (1950): Turing published the paper "Computing Machinery and
Intelligence," introducing the Turing Test as a criterion for machine intelligence.
● Dartmouth Conference (1956): John McCarthy, Marvin Minsky, Nathaniel
Rochester, and Claude Shannon organized this seminal conference, marking the
formal beginning of AI research. The term "Artificial Intelligence" was coined
here.

3. The Golden Years (1950s - 1970s)

● Symbolic AI: Early AI systems, known as symbolic AI or "good old-fashioned AI"
(GOFAI), used rules and logic to simulate human reasoning.
● Expert Systems: The development of systems like DENDRAL (for chemical
analysis) and MYCIN (for medical diagnosis) showcased the potential of AI in
specialized fields.
 ● Machine Learning: Early neural networks, such as the Perceptron (1958), were
introduced, although they faced limitations.

4. Challenges and Setbacks (1970s - 1980s)

● AI Winter: Due to unmet expectations and limited computational power, funding
and interest in AI research waned, leading to periods known as "AI winters."
● Limitations of Early Models: Expert systems struggled with flexibility and
adaptation, and symbolic approaches faced challenges in scaling.

5. Resurgence and Growth (1980s - 1990s)

● Reinvention of Neural Networks: The backpropagation algorithm revitalized
interest in neural networks, allowing for more complex models.
● Advancements in Robotics: AI began to see practical applications in robotics,
with systems capable of basic tasks and navigation.

6. The Era of Big Data and Machine Learning (2000s -

2010s)
● Data-Driven Approaches: The advent of big data and improved computational
power led to breakthroughs in machine learning and deep learning.
● Deep Learning: The success of deep learning models, especially convolutional
neural networks (CNNs) for image recognition, transformed the AI landscape.
Key achievements include:
○ AlexNet (2012): A landmark deep learning model that won the ImageNet
competition, demonstrating the potential of neural networks.

7. AI in Everyday Life (2010s - Present)

● Natural Language Processing: Technologies like chatbots, virtual assistants
(e.g., Siri, Alexa), and language translation systems have become mainstream.
● Reinforcement Learning: Algorithms like AlphaGo (2016) showcased AI's
capability in complex strategic games, defeating world champions.
● Ethical and Societal Considerations: As AI applications grew, concerns about
bias, ethics, and the impact on jobs and privacy emerged, prompting discussions
on responsible AI.
The historical journey of AI reflects a combination of ambitious visions, technical
challenges, and breakthroughs. From early theoretical foundations to the current
era of deep learning and real-world applications, AI continues to evolve, shaping
various aspects of society and industry. Understanding this history provides context
for the ongoing developments and future directions of AI technology.

3.FUTURE OF ARTIFICIAL INTELLIGENCE

1.Enhanced Automation: More tasks will be automated, reducing human error
and increasing efficiency.
2.Natural Language Processing: AI will better understand and generate human
language, improving communication interfaces.
3.AI Ethics and Regulations: As AI becomes pervasive, ethical guidelines and
regulations will be developed to manage its impact.
4.Human-AI Collaboration: Future AI systems will focus on enhancing human
capabilities rather than replacing them.
5.Personalization: AI will continue to create tailored experiences in various fields,
from marketing to education.

Challenges

● Bias in AI: Addressing the biases in AI models is crucial for fair

decision-making.
● Job Displacement: As automation increases, there will be significant shifts
in the job market.
● Security Risks: With advancements come new vulnerabilities and security
threats.

4.INTELLIGENT AGENTS
Intelligent agents are systems that can perceive their environment, reason about it,
and take actions to achieve specific goals. They operate autonomously or
semi-autonomously, making decisions based on their programming and learned
experiences. Here are some key features and types of intelligent agents:

Key Features

1. Perception: Intelligent agents can gather data from their surroundings using
sensors or inputs (e.g., cameras, microphones, or data feeds).
2. Reasoning: They analyze the gathered information to make decisions. This
can involve logic, rules, or machine learning algorithms.
3. Action: Based on their reasoning, intelligent agents can perform actions to
influence their environment (e.g., moving, making recommendations, or
providing information).
4. Learning: Many intelligent agents can improve their performance over time
by learning from past experiences, adapting to new situations.

Typical Intelligent Agents

Examples

1. Virtual Assistants: Systems like Siri and Alexa that understand voice
commands and perform tasks.
2. Self-driving Cars: Use sensors and AI algorithms to navigate and drive
autonomously.
3. Recommendation Systems: Algorithms that suggest products or content
based on user preferences.
4. Game AI: Intelligent agents in video games that adapt to player behavior.

Applications Across Industries

● Healthcare: Diagnostic tools and patient monitoring systems.

● Finance: Fraud detection and algorithmic trading.
● Manufacturing: Robotics and supply chain optimization.
 5.Types of Intelligent Agents
1. Simple Reflex Agents
○ Description: These agents act solely on the current percept, using a
set of condition-action rules (if-then rules). They do not consider the
history of past actions or percepts.
○ Example: A thermostat that adjusts temperature based on current
readings.
2. Model-Based Reflex Agents
○ Description: These agents maintain an internal model of the world,
allowing them to keep track of the state and make decisions based on
the current percept and their model of the environment.
○ Example: A robot vacuum cleaner that uses a map of the room to
navigate and avoid obstacles.
3. Goal-Based Agents
○ Description: These agents act to achieve specific goals. They evaluate
different actions based on how well they help achieve their goals,
allowing for more flexible decision-making.
○ Example: A chess-playing AI that considers various possible moves
to achieve victory.
4. Utility-Based Agents
○ Description: These agents not only pursue goals but also consider the
utility of different actions. They choose actions that maximize their
expected utility, which represents the agent's preferences over
different outcomes.
○ Example: A self-driving car that chooses routes based on traffic
conditions and safety.
5. Learning Agents
○ Description: These agents can learn from their experiences and adapt
their behavior over time. They use feedback from their actions to
improve their performance.
○ Example: A recommendation system that learns user preferences
based on past interactions.
6. Multi-Agent Systems
 ○ Description: Composed of multiple intelligent agents that interact
with each other. These systems can work collaboratively or
competitively to achieve their goals.
○ Example: Autonomous drones coordinating to cover an area for
surveillance.

Intelligent agents are crucial in various applications, from personal assistants to

autonomous systems, enabling machines to interact intelligently with their
environment and make decisions to achieve specific objectives.

Fuzzy Logic
Introduction to Fuzzy Logic
Fuzzy logic is a form of many-valued logic that deals with reasoning that is
approximate rather than fixed and exact. Unlike traditional binary sets (where
variables can be either true or false), fuzzy logic variables may have a truth value
that ranges between 0 and 1. This allows for a more nuanced approach to reasoning
and decision-making, making it particularly useful in systems that require a degree
of tolerance for uncertainty and imprecision.
Key Concepts
1. Fuzzy Sets

● Definition: A fuzzy set is a collection of elements with varying degrees of

membership. Each element has a membership function that quantifies its
degree of belonging to the set.
● Example: In a fuzzy set representing "tall people," someone who is 180 cm
tall might have a membership value of 0.8, while someone who is 160 cm
tall might have a membership value of 0.3.

2. Membership Functions

● Purpose: Membership functions define how each point in the input space is
mapped to a membership value (ranging from 0 to 1).
● Types:
○ Triangular: Defined by three points (the peak and the two base
points).
○ Trapezoidal: Defined by four points, providing a flat top.
○ Gaussian: A bell-shaped curve defined by a mean and standard
deviation.

3. Fuzzy Logic Operations

● Union: The membership value of the union of two fuzzy sets is the
maximum of the membership values.
● Intersection: The membership value of the intersection is the minimum of
the membership values.
● Complement: The complement of a fuzzy set is calculated as 1−μ(x)1 -
\mu(x)1−μ(x), where μ(x)\mu(x)μ(x) is the membership function.

4. Fuzzy Rules

Fuzzy logic systems use a set of rules that express the relationship between input
and output using linguistic variables. For example:

● Rule Example: If temperature is "high," then fan speed is "fast."

 ● These rules can be combined to make complex decisions based on multiple
inputs.

Applications of Fuzzy Logic

Fuzzy logic has a wide range of applications across various fields, including:

1. Control Systems: Used in appliances like washing machines, air

conditioners, and cameras to automate operations based on imprecise input
data.
2. Expert Systems: In medical diagnosis and financial forecasting, fuzzy logic
helps handle uncertain and imprecise information.
3. Decision-Making Systems: Fuzzy logic is utilized in systems that require
human-like reasoning, such as customer satisfaction assessment and risk
analysis.
4. Robotics: Helps robots make decisions in uncertain environments,
improving navigation and task execution.

Advantages of Fuzzy Logic

● Tolerance for Imprecision: Fuzzy logic systems can handle uncertainties
and approximations, making them robust in real-world applications.
● Flexibility: Easy to incorporate human expertise and reasoning into system
design using linguistic variables.
● Simplicity: Often simpler to design than traditional control systems, as it
mimics human reasoning.

Intelligent agents are crucial in various applications, from personal assistants to

autonomous systems, enabling machines to interact intelligently with their
environment and make decisions to achieve specific objectives.

DIFFERENCE BETWEEN FUZZY LOGIC AND CRISP LOGIC

Fuzzy logic and crisp logic (also known as binary or classical logic) are two
different approaches to reasoning and decision-making. Here’s a breakdown of
their key differences:

1. Definition

● Crisp Logic: Operates on binary values—true or false (1 or 0). It follows

strict rules of classical logic where statements are either completely true or
completely false.
● Fuzzy Logic: Allows for degrees of truth. Instead of binary outcomes, it
represents uncertainty and vagueness using values between 0 and 1,
reflecting the concept of partial truth.

2. Truth Values

● Crisp Logic: Only two truth values (true/false). For example, a statement
like "It is raining" can only be true or false.
● Fuzzy Logic: Can have an infinite range of truth values between 0 and 1.
For example, "It is warm" could be represented as 0.7, indicating a degree of
truth.

3. Reasoning Approach

● Crisp Logic: Uses precise and deterministic reasoning. It follows strict rules
and is suited for clear-cut problems.
● Fuzzy Logic: Employs approximate reasoning, which is useful in situations
with uncertainty, imprecision, or vagueness.

4. Applications

● Crisp Logic: Common in traditional computing, digital circuits, and

algorithms requiring exact inputs, such as database queries.
● Fuzzy Logic: Often used in control systems, artificial intelligence, and areas
where human-like reasoning is beneficial, such as in consumer electronics
(e.g., washing machines, air conditioning).
5. Examples

● Crisp Logic: A thermostat that turns on heating when the temperature is

below a certain threshold (e.g., 20°C) and off when it's above.
● Fuzzy Logic: A thermostat that uses fuzzy rules, like “If the temperature is
slightly below 20°C, turn on the heater a little” or “If the temperature is very
low, turn it on fully.”

Crisp logic is effective for clear-cut scenarios, while fuzzy logic provides a
more flexible approach to handle uncertainty and complexity in real-world
situations. Each has its own strengths and applications, making them suitable
for different types of problems.

Fuzzy logic is useful for commercial and practical purposes.

​ It can control machines and consumer products.

​ It may not give accurate reasoning, but acceptable reasoning.
​ Fuzzy logic helps to deal with the uncertainty in engineering.

Fuzzy Logic Systems Architecture

It has four main parts as shown −

It has four main parts as shown −

1.Fuzzification Module − It transforms the system inputs, which are crisp

numbers, into fuzzy sets. It splits the input signal into five steps such as −

LP x is Large Positive

MP x is Medium Positive

S x is Small

MN x is Medium Negative

LN x is Large Negative

2.Knowledge Base − It stores IF-THEN rules provided by experts.

3.Inference Engine − It simulates the human reasoning process by making
fuzzy inference on the inputs and IF-THEN rules.

4.Defuzzification Module − It transforms the fuzzy set obtained by the

inference engine into a crisp value.

Example of a Fuzzy Logic System

Let us consider an air conditioning system with a 5-level fuzzy logic system.
This system adjusts the temperature of the air conditioner by comparing the
room temperature and the target temperature value.
Algorithm

​ 1.Define linguistic Variables and terms (start)

​ 2.Construct membership functions for them. (start)
​ 3.Construct knowledge base of rules (start)
​ 4.Convert crisp data into fuzzy data sets using membership functions.
(fuzzification)
​ 5.Evaluate rules in the rule base. (Inference Engine)
​ 6.Combine results from each rule. (Inference Engine)
​ 7.Convert output data into non-fuzzy values. (defuzzification)
MEMBERSHIP FUNCTION FEATURES:
The different types of environments in AI:

1. Fully Observable
2. Partially Observable
3. Deterministic
4. Stochastic
5. Static
6. Dynamic
7. Discrete
8. Continuous
9. Single-Agent
10.Multi-Agent
11.Known
12.Unknown
13.Physical
14.Abstract

Fuzzy If-Then Rules

Fuzzy If-Then rules are a fundamental component of fuzzy logic systems, allowing
for the representation of knowledge and decision-making in a way that mimics
human reasoning. These rules enable systems to handle imprecise and uncertain
information.

Structure of Fuzzy If-Then Rules

A typical fuzzy If-Then rule has the following structure:
If (condition) Then (conclusion)

Components:

1. Antecedent (Condition): This part of the rule specifies the criteria that need
to be satisfied. It often involves fuzzy sets and linguistic variables.
○ Example: "If the temperature is high"
2. Consequent (Conclusion): This part indicates the action or outcome that
follows if the antecedent is true.
○ Example: "Then the fan speed is fast"

Example of a Fuzzy Rule

● Rule: If the temperature is "high" Then the fan speed is "fast."

Using Linguistic Variables

Fuzzy rules often use linguistic terms to represent conditions and conclusions.
Common linguistic variables include:

● Temperature: low, medium, high

● Speed: slow, medium, fast

Fuzzy Rule Example Set

Here’s a set of fuzzy If-Then rules for a simple climate control system:

1. Rule 1: If the temperature is "low" Then the heater is "high."

2. Rule 2: If the temperature is "medium" Then the heater is "medium."
3. Rule 3: If the temperature is "high" Then the heater is "off."

Fuzzy Inference System (FIS)

Fuzzy If-Then rules are used within a Fuzzy Inference System (FIS) to make
decisions based on input data. The FIS typically consists of three main stages:

1. Fuzzification: Converts crisp input values into fuzzy values using

membership functions.
2. Rule Evaluation: Applies fuzzy If-Then rules to the fuzzy input data.
3. Defuzzification: Converts fuzzy output values back into crisp values to
provide a decision or action.
 FUZZY RELATION
A fuzzy relation is a mathematical concept used in fuzzy set theory that generalizes traditional
relations by allowing for degrees of membership rather than just binary outcomes (true/false or
0/1). In a fuzzy relation, elements can have varying degrees of association.

For example, in a traditional relation between sets AAA and BBB, an element a∈Aa \in Aa∈A
is either related to b∈Bb \in Bb∈B or it is not. In a fuzzy relation, the relationship can be
expressed with a membership value between 0 and 1, where:

● 0 means no relationship,
● 1 means a full relationship,
● values in between indicate partial relationships.

Fuzzy relations can be used in various applications, such as decision-making, control systems,
and pattern recognition, where uncertainty and vagueness are present. They help to model
complex systems where binary logic falls short, providing a more nuanced understanding of
relationships among elements.

A fuzzy relation is the cartesian product of mathematical fuzzy sets. Two fuzzy sets are
taken as input, the fuzzy relation is then equal to the cross product of the sets which is
created by vector multiplication. Usually, a rule base is stored in a matrix notation which
allows the fuzzy controller to update its internal values.

From a historical perspective, the first fuzzy relation was mentioned in the year 1971 by
Lotfi A. Zadeh.

A practical approach to describe a fuzzy relation is based on a 2d table. At first, a table

is created which consists of fuzzy values from 0..1. The next step is to apply the
if-then-rules to the values. The resulting numbers are stored in the table as an array.

Fuzzy relations can be utilized in fuzzy databases.

Cartesian Product
6.PROBLEM SOLVING APPROACH TO TYPICAL
INTELLIGENT PROBLEMS

When tackling typical AI problems, a structured problem-solving approach can be

highly effective. Here’s a step-by-step framework you can follow:

1. Define the Problem

● Clarify the Goal: Understand what you want to achieve.

● Identify Constraints: Recognize any limitations (e.g., data availability,
computational resources).

2. Gather Information

● Collect Data: Determine the type and amount of data needed.

● Review Literature: Look at existing solutions and relevant research.

3. Preprocess Data

● Clean the Data: Handle missing values, outliers, and noise.

● Transform the Data: Normalize, standardize, or encode categorical
variables as necessary.

4. Choose the Right Model

● Select Algorithms: Depending on the problem type (classification,

regression, clustering, etc.), choose appropriate models.
● Consider Hybrid Approaches: Sometimes combining models can yield
better results.

5. Train the Model

● Split the Data: Use training, validation, and test sets to evaluate model
performance.
● Tune Hyperparameters: Optimize the model settings using techniques like
grid search or random search.

6. Evaluate the Model

● Use Metrics: Assess model performance using metrics relevant to the

problem (accuracy, precision, recall, F1-score, etc.).
● Analyze Results: Look for patterns in errors to identify areas for
improvement.

7. Iterate and Improve

● Refine Features: Feature engineering can significantly enhance model

performance.
● Try Different Models: Experiment with different algorithms or ensembles.

8. Deploy the Solution

● Prepare for Production: Ensure the model is robust, scalable, and secure.
● Monitor Performance: Implement monitoring to catch issues in real-time.

9. Maintain and Update

● Retrain as Necessary: Periodically update the model with new data to

maintain accuracy.
● Incorporate Feedback: Use user feedback to refine the system.

Example Application: Image Classification

1. Define the Problem: Classify images of animals into categories (dogs, cats,
etc.).
2. Gather Information: Collect a dataset of labeled images.
3. Preprocess Data: Resize images, augment data, and normalize pixel values.
4. Choose the Right Model: Select a convolutional neural network (CNN) for
image data.
5. Train the Model: Split data into training, validation, and test sets; tune the
model.
 6. Evaluate the Model: Use accuracy and confusion matrix to assess
performance.
7. Iterate and Improve: Experiment with different CNN architectures and data
augmentation techniques.
8. Deploy the Solution: Set up the model in a web application for users to
upload images.
9. Maintain and Update: Monitor user inputs and regularly update the model
with new labeled images.

By following this structured approach, you can systematically tackle AI problems

and optimize your solutions effectively.

DIFFERENCE BETWEEN INFORMED AND

UNINFORMED SEARCH TECHNIQUES

Parameters Informed Search Uninformed Search

It is also known as It is also known as

Known as
Heuristic Search. Blind Search.

It doesn’t use
Using It uses knowledge for
knowledge for the
Knowledge the searching process.
searching process.
 It finds solution slow
It finds a solution more
Performance as compared to an
quickly.
informed search.

It may or may not be

Completion It is always complete.
complete.

Cost Factor Cost is low. Cost is high.

It consumes less time It consumes moderate

Time because of quick time because of slow
searching. searching.

No suggestion is given
There is a direction given
Direction regarding the solution
about the solution.
in it.

It is less lengthy while It is more lengthy

Implementation
implemented. while implemented.
 It is more efficient as
It is comparatively less
efficiency takes into
efficient as incurred
account cost and
cost is more and the
Efficiency performance. The
speed of finding the
incurred cost is less and
Breadth-Firstsolution
speed of finding
is slow.
solutions is quick.

Computational Comparatively higher

Computational
requirements are computational
requirements
lessened. requirements.

Having a wide scope in Solving a massive

Size of search
terms of handling large search task is
problems
search problems. challenging.