R. C.

Patel Institute of Technology, Shirpur

Fundamentals of Artificial
Intelligence & Data Science
(RCP23FCPC2AD)
FIRST YEAR B.TECH (ALL BRANCHES) WITH EFFECT FROM
YEAR 2023-24
A.Y.2024-25, SEM-II

PROF.S.L.TAMBE
 Subject Structure

RCPIT-Shirpur Fundamentals ofAIDS

 Syllabus
Unit-I Introduction to Artificial Intelligence 06 Hrs.
Introduction, History of Artificial Intelligence, Types of AI, Intelligent Systems:
Categorization of Intelligent System, Components of AI Program, Applications of
AI (Robotics, Healthcare, Industry), Current trends in AI.
Intelligent Agents: Agents and Environments, The concept of rationality, The
nature of environment, PEAS representation, The structure of Agents, Types of
Agents, and Learning Agent.

Unit-II Problem solving and Searching Techniques 06 Hrs.

Problem Solving Agent, Formulating Problems: Steps problem solving in AI
Example Problems: Vacuum world state space graph, 8 Puzzle Problem, 4
queen and 8 Queen Problem. Introduction to searching techniques in AI.

RCPIT-Shirpur FAIDS UNIT-01

 Syllabus
Unit-III Expert Systems 06 Hrs.
Introduction, Phases in building Expert Systems, Expert system Architecture, Case
Study on Expert System. Applications of Expert Systems. How to achieve AI in
practical: Introduction to Machine and Deep Learning, Introduction to No code AI
tools.
Unit-IV Fundamentals of Data Science 08 Hrs.
Introduction to data lake, Data Lake frameworks, Data Pre-Processing An
Overview, Cleaning, Data Integration, Data Reduction, Data Transformation and
Data discretization. Exploratory Data Analysis (EDA): Philosophy of EDA - The Data
Science Process
 Unit-01

Introduction to Artificial Intelligence

 Artificial intelligence (AI)
Artificial intelligence (AI),

In its broadest sense, is intelligence exhibited by machines, particularly

computer systems. It is a field of research in computer science that
develops and studies methods and software that enable machines to
perceive their environment and use learning and intelligence to take
actions that maximize their chances of achieving defined goals. Such
machines may be called AIs.
https://www.youtube.com/watch?v=VIj2f43sVS4
(You Tube Link)
 AI History
The field of AI research was founded at a workshop held on the campus of
Dartmouth College in 1956.[1] Attendees of the workshop became the leaders of
AI research for decades. Many of them predicted that machines as intelligent as
humans would exist within a generation.
The Dawn of Artificial Intelligence (1950s-1960s)
•1950 : In 1950 saw the publication of Alan Turing's work ,"Computing Machinery
and Intelligence " which introduced the Turing Test—a measure of computer
intelligence.

•1956: A significant turning point in AI research occurs in 1956 when, John

McCarthy first uses the phrase "Artificial Intelligence" at the Dartmouth
Workshop.
 AI History
•1950s–1960s: The goal of early artificial intelligence (AI) research was to encode
human knowledge into computer programs through the use of symbolic
reasoning, and logic-based environments.50 : In 1950 saw the publication of Alan
Turing's work ,"Computing Machinery and Intelligence " which introduced the
Turing Test—a measure of computer intelligence.

•1956: A significant turning point in AI research occurs in 1956 when, John

McCarthy first uses the phrase "Artificial Intelligence" at the Dartmouth
Workshop.

•1950s–1960s: The goal of early artificial intelligence (AI) research was to encode
human knowledge into computer programs through the use of symbolic
reasoning, and logic-based environments.
 AI History
AI’s Early Achievements and Setbacks (1970s-1980s)
This age has seen notable developments as well as difficulties :
1970: The 1970s witnessed the development of expert systems , which were
intended to capture the knowledge of experts in a variety of domains. Data
Scientists created rule-based systems that , could use pre-established guidelines
to address certain issues.
Limitations: Due to their inability to handle ambiguity and complicated
circumstances , these systems had a limited range of applications.
The Artificial Intelligence Winter(1970–1980): A period of inactivity brought on by
a lack of funding , and un-met expectations.
 AI History
The 1990s bring a transformative move in AI :
1990s: A worldview move towards machine learning approaches happens.
• Rise of Machine-Learning
• Neural Organize Insurgency
• Information Powers AI
• Unused Areas Rise
• Brilliant Age of AI
• Improved Execution
 AI History
The AI Boom: Deep Learning and Neural Networks (2000s-2010s)
The 21st century , witnesses the rise of profound learning , and neural systems :
2000s-2010s: Profound learning a subset of machine learning imitating the
human brain's structure and work , came to the cutting edge.
• Profound Neural Systems
• Innovative Progressions
• Corporate Speculation
• Counterfeit Neural Systems
 Generative Pre-trained Transformers: A New Era (GPT Series)
A novel advancement in recent times is the use of Generative Pre-trained Transformers :
GPT Series: Trained on enormous volumes of textual data , these models have rocked the globe.
GPT-3: This model transforms language processing by producing writing that is similar to that of
a human being and translating between languages.
Learning from Text: Large volumes of text are absorbed by GPT models, such as - GPT-3, which
help them comprehend syntax, context , and comedy.
Beyond Translation: GPT-3 serves as a portable writing assistant by producing essays, poetry ,
and even language translations.
The Upcoming Generation: This new wave of models , which can write, translate and generate
original material as well as provide insightful responses, is exemplified by models such as Bard,
ChatGPT, and Bing Copilot.
Pushing Boundaries: These developments have increased the possible applications of AI
showcasing its ability in content production, creative projects and language translation.
Based on Capabilities:
1) Narrow AI
Also called weak AI, this type of AI is designed to perform specific tasks.
It can't learn independently, but follows rules or learns patterns from data.
Examples include voice assistants, recommendation engines, and image
recognition systems.
2) General AI
Also called artificial general intelligence (AGI), this type of AI is designed to think,
reason, and act like humans. It can learn, reason, and make decisions in many
contexts. AGI is still theoretical and has not yet been achieved.
3) Super AI (Artificial Superintelligence)
• Capable of independent thinking, reasoning, and problem-solving at a superior level.
• Artificial Super Intelligence (ASI) is really good at solving hard problems, like stopping climate
change or finding treatments for illnesses.

•ASI could help us create new things much faster in areas like medicine, energy, and exploring
space.

•ASI could help scientists find things quicker, like learning more about space, genes, or new
materials.

•ASI might help countries cooperate better by giving good advice and helping make decisions
that benefit everyone.

•ASI could help people communicate better, even if they speak different languages or come from
different cultures.
Based on Functionalities :
1) Reactive machines
These AI systems have no memory and are task-specific.
They can copy how the human mind reacts to stimulus (enjoyment).
Examples include machine learning models that use customer data to make
recommendations.
2) Self-aware AI
This type of AI would have consciousness and be able to understand its own
emotions and traits(a quality that forms part of your character or personality).
It would also have its own emotions, needs, and beliefs.
Self-aware AI is still theoretical and has not yet been achieved.
3) Limited Memory AI :
Definition:
Limited Memory AI refers to AI systems that can learn from past data and make
decisions based on historical information. Unlike Reactive AI (which has no
memory), these systems can improve over time by analyzing previous
experiences.
Key Features:
• Can store and use past data for better decision-making.
• Common in machine learning & deep learning models.
• Used in self-driving cars, chatbots, recommendation systems, and fraud
detection.
4) Theory of Mind AI :
Definition:
Theory of Mind AI refers to a hypothetical(based on a possible idea) type of
artificial intelligence that can understand human emotions, beliefs,
intentions, and thoughts. Unlike current AI, which processes data without
true comprehension, this AI would be capable of social intelligence and
human-like interactions.
 Intelligent System in AI
What are Intelligent System?
• Intelligent systems in artificial intelligence (AI) represent a broad class of systems
equipped with algorithms that can perform tasks typically requiring human
intelligence.
• These systems span various domains from robotics to data analysis, playing a
pivotal role in driving innovation across industries.
• An intelligent system in AI is a technology equipped with the capability to gather
data, process it, and make decisions or perform actions based on that data.
• At its core, an intelligent system mimics the cognitive functions of human beings,
such as learning from experience, understanding complex concepts, solving
problems, and making decisions.
 Categorization of Intelligent System
Intelligent systems in AI can be categorized in various ways based on functionality, learning
capability, autonomy, and application domains.
1. Based on Capability
a) Narrow AI (Weak AI)
•Designed for specific tasks.
•Cannot perform beyond its programming.
•Examples: Chatbots, recommendation systems, image recognition.
b) General AI (Strong AI)
•Mimics human intelligence across various tasks.
•Can learn and adapt without human intervention.
•Currently theoretical (not yet achieved).
 Categorization of Intelligent System
c) Super AI
•Hypothetical AI surpassing human intelligence.
•Potential for self-awareness and decision-making.
•Exists only in theoretical discussions.
2. Based on Learning Approach
a) Supervised Learning Systems
•Learn from labeled data.
•Examples: Spam filters, facial recognition.
b) Unsupervised Learning Systems
•Identify patterns in unlabeled data.
•Examples: Customer segmentation, anomaly detection.
 Categorization of Intelligent System
c) Reinforcement Learning Systems
•Learn through trial and error with rewards.
•Examples: AlphaGo, autonomous robots.
d) Self-Learning Systems
•Capable of adjusting without predefined models.
•Examples: Advanced AI personal assistants.
3. Based on Autonomy
a) Reactive Machines
b) Limited Memory Systems
c) Theory of Mind AI
d) Self-Aware AI
 Categorization of Intelligent System
4. Based on Functionality
a) Expert Systems
•Mimic human experts in a specific domain.
•Example: Medical diagnosis AI.
b) Neural Networks
•Mimic human brain functions.
•Example: Deep learning applications.
c) Fuzzy Logic Systems
•Handle uncertainty and approximate reasoning.
•Example: AI in automatic climate control.
d) Hybrid AI Systems
•Combine multiple AI techniques.
•Example: Chatbots using NLP + deep learning.
 The components of an artificial intelligence (AI) program
• Natural language processing (NLP): The ability to understand and respond to human
language .
• Machine learning (ML): The ability for machines to learn from data.
• Computer vision: The use of machine learning and neural networks to teach computers to see
.
• Deep learning: A method of AI that teaches computers to process data in a way similar to the
human brain .
• Problem solving: The ability to use techniques like planning, search, and optimization to solve
problems .
• Expert systems: The use of knowledge bases to solve complex problems.
• Perception: The use of real or simulated sensory organs to interpret data and identify objects.
•Neural networks: A method of AI that teaches computers to process data in a way similar to
the human brain.
These components work together to enable machines to perform complex tasks and interact with the world.
 Applications of AI (Robotics, Healthcare, Industry)
1. Robotics
•Autonomous Robots – Self-navigating robots in warehouses (e.g., Amazon's Kiva
robots).
•Humanoid Robots – AI-powered robots for assistance (e.g., Sophia by Hanson
Robotics).
•Military & Defense – AI-powered drones and surveillance robots.
•Agricultural Robots – Automated crop monitoring and harvesting.
•Disaster Response – AI-driven robots for rescue operations.
Applications of AI (Robotics, Healthcare, Industry)

2. Healthcare
• Medical Diagnosis – AI detects diseases from medical images (e.g., IBM Watson,
Google DeepMind).
• Personalized Treatment – AI recommends treatment plans based on patient
history.
• Surgical Robots – AI-assisted robotic surgery (e.g., Da Vinci Surgical System).
• Drug Discovery – AI accelerates new drug development.
• Healthcare Chatbots – Virtual assistants help patients with medical queries.
Applications of AI (Robotics, Healthcare, Industry)
3. Industry & Manufacturing
•Predictive Maintenance – AI predicts machinery failures before they occur.
•Quality Control – AI-powered visual inspection for defect detection.
•Supply Chain Optimization – AI enhances logistics and inventory management.
•Smart Factories – AI-powered automation in Industry 4.0.
•Energy Efficiency – AI optimizes energy consumption in manufacturing plants.
Current trends in AI
 Current trends in AI
1. Generative AI
• AI models like ChatGPT and DALL·E generate text, images, and videos.
• Used in content creation, coding, and design.
2. AI in Healthcare
• AI-powered drug discovery and personalized medicine.
• AI-assisted diagnostics using medical imaging.
3. AI in Automation & Robotics
• Increased use of AI-driven robots in industries and homes.
• AI-powered drones for delivery and surveillance.
 Current trends in AI
4. Ethical AI & Regulation
• Governments implementing AI policies and ethical guidelines.
• Focus on bias reduction and responsible AI.
5. AI in Cybersecurity
• AI detects fraud, phishing attacks, and security threats.
• Automated threat response systems.
6. AI for Climate & Sustainability
• AI optimizes energy consumption and predicts climate changes.
• Used in smart grids and carbon footprint reduction.
 Current trends in AI
7. Multimodal AI
• AI models process multiple data types (text, image, audio, video).
• Example: OpenAI’s GPT-4 Turbo with vision capabilities.
8. AI in Finance & Business
• AI-driven stock market predictions.
• Chatbots and virtual assistants in banking.
9. AI & Quantum Computing
• Research on using AI with quantum computing for faster problem-solving.
10. AI-Powered Edge Computing
• AI models running on devices instead of cloud-based systems (e.g.,
smartphones, IoT devices).
 Intelligent Agents and Environments
1. Intelligent Agents:
An Intelligent Agent (IA) is an entity in Artificial Intelligence (AI) that
perceives its environment and takes actions to achieve specific goals.
These agents can be software-based (like chatbots) or hardware-based
(like robots).

1. Agents in AI
An agent is anything that can sense its
environment through sensors and act
upon it using actuators.
 Intelligent Agents
Types of Intelligent Agents:
a) Simple Reflex Agents
•Act only based on current perception.
•No memory of past actions.
•Example: Thermostat controlling room temperature.
b) Model-Based Reflex Agents
•Maintain an internal model of the environment.
•Use past perceptions to make better decisions.
•Example: Self-driving cars considering past road conditions.
 Intelligent Agents
Types of Intelligent Agents:
c) Goal-Based Agents
•Make decisions based on defined goals.
•Use search and planning algorithms.
•Example: AI in GPS navigation systems.
d) Utility-Based Agents
•Optimize for the best outcome using a utility function.
•Not just goal-oriented but also efficiency-driven.
•Example: AI in financial trading maximizing profits.
 Environments Agents
2. Environments in AI
An environment is the world in which an AI agent operates. It provides inputs
(perceptions) that the agent uses to make decisions.
Types of Environments:
a) Fully Observable vs. Partially Observable
•Fully Observable: Agent has complete knowledge (e.g., Chess game).
•Partially Observable: Agent has limited perception (e.g., Poker game).
b) Deterministic vs. Stochastic
•Deterministic: Outcome is predictable (e.g., Tic-Tac-Toe).
•Stochastic: Outcome has randomness (e.g., Stock Market).
 Environments Agents
2. Environments in AI
c) Static vs. Dynamic
•Static: Environment does not change (e.g., Crossword puzzles).
•Dynamic: Environment changes with time (e.g., Autonomous driving).
d) Discrete vs. Continuous
•Discrete: Limited number of possible actions (e.g., Chess).
•Continuous: Infinite possibilities (e.g., Robotics movement).
e) Single-Agent vs. Multi-Agent
•Single-Agent: One agent acts alone (e.g., Sudoku solver).
•Multi-Agent: Multiple agents interact (e.g., Online multiplayer games).
 The Concept of Rationality in AI
The Concept of Rationality in AI
• Rationality in AI refers to an agent’s ability to make the best possible decision based
on its knowledge, perception, and goals.
• A rational agent is one that acts optimally to maximize its performance measure in a
given environment.
"acts optimally" means making the best possible decision based on the available information, resources, and goals

Key Aspects of Rationality:

1. Performance Measure – Defines what is considered a "good" action.
2. Percept Sequence – The history of inputs an agent receives.
3. Knowledge of the Environment – How much the agent knows about its
surroundings.
4. Possible Actions – The choices available to the agent.
 The Concept of Rationality in AI
Types of Rationality in AI:
1. Perfect Rationality – Theoretical ideal where an agent always makes the best choice (often
impossible in real-world scenarios).
2. Bounded Rationality – Practical approach where agents make the best decision within
computational limits.
3. Utility-Based Rationality – Agents choose actions that maximize a utility function (e.g., cost
vs. reward).
Example:
Self-Driving Car:
• A rational AI should follow traffic rules, avoid collisions, and reach the destination efficiently
based on real-time sensor data and traffic conditions.
 PEAS representation
PEAS Representation in AI
The PEAS framework is used to define an AI agent's environment and how it
interacts with it. It helps design intelligent agents by breaking down their
functionality into four key components:
PEAS Components
P (Performance Measure) – The criteria that determine an agent’s success.
E (Environment) – The surroundings in which the agent operates.
A (Actuators) – The mechanisms the agent uses to take action.
S (Sensors) – The devices or methods the agent uses to perceive the environment.
 PEAS representation
Examples of PEAS for Different AI Systems
1. Self-Driving Car
• Performance Measure: Safety, fuel efficiency, travel time, obeying traffic laws.
• Environment: Roads, traffic, pedestrians, weather conditions.
• Actuators: Steering wheel, brakes, accelerator, turn signals.
• Sensors: Cameras, LiDAR, GPS, speedometer.
2. Smart Vacuum Cleaner
• Performance Measure: Cleanliness, battery efficiency, time taken.
• Environment: Floor, furniture, obstacles, dust.
• Actuators: Wheels, brushes, vacuum motor.
• Sensors: Dirt sensors, bump sensors, cameras