AI Prof Deepali Virmani

What is Artificial Intelligence?

Artificial intelligence (AI) is technology that enables

computers and machines to simulate human learning,
comprehension, problem solving, decision making, creativity
and autonomy.
Artificial intelligence (AI) is the simulation of human
intelligence in machines that are programmed to think and
act like humans. Learning, reasoning, problem-solving,
perception, and language comprehension are all examples of
cognitive abilities.

Artificial Intelligence is a method of making a computer, a

computer-controlled robot, or a software think intelligently
like the human mind. AI is accomplished by studying the
patterns of the human brain and by analyzing the cognitive
process. The outcome of these studies develops intelligent
software and systems.
For example, Natural Language Processing (NLP) uses AI to analyze
and interpret human language in text or speech. It enables
applications like chatbots and virtual assistants to understand user
queries, extract meaningful information, and deliver accurate,
context-aware responses, transforming unstructured communication
into actionable insights.
1. Based on Capabilities
2. Based on Functionalities
- Reactive Machines
These AI systems do not store memories or past experiences for future actions. They analyze and
respond to different situations. IBM's Deep Blue, which beat Garry Kasparov at chess, is an
example.

- Limited Memory
These AI systems can make informed and improved decisions by studying the past data they have
collected. Most present-day AI applications, from chatbots and virtual assistants to self-driving
cars, fall into this category.

- Theory of Mind
This is a more advanced type of AI that researchers are still working on. It would entail
understanding and remembering emotions, beliefs, needs, and depending on those, making
decisions. This type requires the machine to understand humans truly.

- Self-aware AI
This represents the future of AI, where machines will have their own consciousness, sentience,
and self-awareness. This type of AI is still theoretical and would be capable of understanding and
possessing emotions, which could lead them to form beliefs and desires.
3. Based on Technologies

- Machine Learning (ML)

AI systems capable of self-improvement through experience, without direct programming. They concentrate on
creating software that can independently learn by accessing and utilizing data.

- Deep Learning
A subset of ML involving many layers of neural networks. It is used for learning from large amounts of data and is the
technology behind voice control in consumer devices, image recognition, and many other applications.

- Natural Language Processing (NLP)

This AI technology enables machines to understand and interpret human language. It's used in chatbots, translation
services, and sentiment analysis applications.

- Robotics
This field involves designing, constructing, operating, and using robots and computer systems for controlling them,
sensory feedback, and information processing.

- Computer Vision
This technology allows machines to interpret the world visually, and it's used in various applications such as medical
image analysis, surveillance, and manufacturing.

- Expert Systems
These AI systems answer questions and solve problems in a specific domain of expertise using rule-based systems.
https://www.youtube.com/watch?v=VNz3KGoAhG4&t=363s
Advantages of Artificial
Intelligence
 The Foundation of AI: Core Principles and Components
1. Data and Knowledge Representation

The foundation of artificial intelligence is data. An AI system needs huge amounts of data to find patterns and acquire insights to "learn" and make decisions. The goal of
knowledge representation, on the other hand, is to organize this material such that computers can understand, store, and retrieve it. The core of artificial intelligence is
knowledge and data, which work together to allow it to understand unprocessed data.

2. Algorithms

An AI system can assess data and draw conclusions from it because of algorithms, which are detailed instructions or sets of rules. These algorithms have become more
and more advanced, enabling AI to perform incredibly difficult jobs like language translation and image identification. These algorithms provide the basis of artificial
intelligence, allowing it to learn from data and make predictions.

3. Statistics and Mathematics

The foundation for understanding and evaluating data patterns is mathematics, especially statistics and calculus. Statistical techniques help in the development of models
that show connections in data, enabling AI to make sensible choices. AI would have very little predictive capacity without mathematics.

4. Neuroscience and Cognitive Science

Researchers have learned how to reproduce specific cognitive processes in AI systems by examining the human brain. By providing a blueprint of human learning,
memory, and decision-making, cognitive science and neuroscience help build more "human-like" robots, which form the basis of artificial intelligence.

5. Philosophy and Ethics

In AI, ethical issues are becoming more and more important. By establishing guidelines for how AI should function and be used in society, concerns about privacy, biases,
and responsibility help to build the field's foundation. Ethical standards ensure that advancements in AI protect individual rights and are consistent with the values of
society.
 AI Techniques
The key concepts of AI development revolve around creating intelligent machines that can perform tasks typically done by
humans, such as problem-solving, decision-making, and learning. These machines require complex algorithms and large
amounts of data to learn from and improve their performance over time.

Data Collection and Preprocessing

In the realm of AI development, the significance of high-quality data cannot be overstated. At the very foundation of creating
intelligent systems that can learn, predict, and make informed decisions is the collection and preprocessing of data.

Collecting data is the first critical step, which involves gathering relevant information from various sources that can be used to
train machine learning models. This data must then undergo a crucial phase known as preprocessing, where it is cleaned,
normalized, and transformed to ensure consistency and accuracy.

By meticulously collecting and preprocessing data, developers can lay a solid foundation for AI systems, enabling them to
operate more efficiently and effectively. This process not only enhances the model's performance but also significantly reduces
the chances of biases and errors, leading to more reliable and trustworthy AI solutions.

Machine Learning Algorithms

Understanding the different machine learning algorithms is pivotal in AI development for creating intelligent systems that can
learn from data and improve over time. First, we have supervised learning, where the algorithm learns from labeled data,
making predictions or decisions based on input-output pairs. It's like teaching a child with examples.

Next, unsupervised learning involves training an algorithm on data without explicit instructions, allowing it to identify patterns
and relationships on its own—a bit like solving a puzzle without the picture on the box.

Lastly, reinforcement learning takes a different approach by rewarding the algorithm for making correct decisions and
penalizing it for errors, akin to training a pet with treats and corrections. Each of these learning algorithms plays a crucial role
in the development of AI, enabling machines to tackle complex tasks, from voice recognition to autonomous driving.
Deep Learning Techniques

Deep learning is revolutionizing AI development through sophisticated techniques, notably neural networks. At the heart of this
innovation are Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs).

CNNs are particularly adept at processing visual information, making them indispensable for tasks like image and video
recognition. They simulate how the human brain processes visual data, identifying patterns and features with remarkable
accuracy.

RNNs, on the other hand, excel in handling sequential data, such as text or speech. Their architecture allows them to remember
information from previous inputs, making them ideal for language translation, sentiment analysis, and speech recognition.
Together, CNNs and RNNs form the backbone of deep learning techniques, pushing the boundaries of what AI can achieve.

Natural Language Processing

Natural Language Processing (NLP) allows machines to comprehend, interpret, and generate human language. This AI technique
has paved the way for virtual assistants, chatbots, and language translation tools, making communication between humans and
machines more seamless than ever.

Computer Vision

Computer Vision equips machines with the ability to interpret visual information from the world. This technique has revolutionized
industries like healthcare, automotive, and robotics, enabling tasks such as facial recognition, object detection, and autonomous
driving.
 Defining a problem as a state space search

Defining a problem as a state space search is a technique in artificial intelligence (AI) that
represents a problem as a set of states and the transitions between them. The goal is to find a
path from the initial state to the desired goal. This approach can be used to solve a variety of AI
issues, including pathfinding, solving puzzles, playing games, and more. The fundamental
concept is to visualize the issue as a graph with nodes standing in for states and edges for
transitions.
Important ideas consist of:
•State: A specific configuration of the problem.
•Initial State: The starting point of the search.
•Goal State: The desired end configuration.
•Transition: An action that changes one state to another.
•Path: A sequence of states connected by transitions.
•Search Strategy: The method used to explore the state
space.
What is a Production System in AI?

A production system, also known as a rule-based system, is a type of artificial

intelligence software designed to mimic the problem-solving ability of human

experts. It consists of a knowledge base of rules and an inference engine that

applies those rules to solve problems or make decisions within a specific domain.
The components of Production System in AI encompass three essential
elements:

Global Database / Working Memory: Also called the global database,

this is a temporary storage area that holds facts about the current state
of the problem or situation being analyzed by the system.

Production Rules / Knowledge Base: This is a collection of rules that

encode domain-specific knowledge. Rules typically take the form of “IF
(condition) THEN (action)”. For example, an expert system for medical
diagnosis might have a rule like “IF the patient has a fever AND a rash,
THEN there is a possibility of measles.”

Control System / Inference Engine: This is the control mechanism that

iteratively evaluates the rules from the knowledge base against the
contents of the working memory. It determines which rules are applicable
and fires (executes) them, updating the working memory with new facts
derived from applying the rules.
How a Production System Works:
•The working memory is initialized with the known facts about the problem.
•The inference engine matches the rules in the knowledge base against the facts in working memory.
•Rules whose conditions are satisfied are placed on the agenda.
•A conflict resolution strategy (e.g. priority, recency, etc.) is used to select one rule from the agenda to fire.
•The selected rule is fired, and its actions update the working memory.
•The cycle repeats from step 2 until a solution is found or no more rules can be fired.
 Simplicity

The production rule in AI is in the form of an ‘IF-THEN’ statement.

Every rule in the production system has a unique structure. It helps
represent knowledge and reasoning in the simplest way possible to
solve real-world problems. Also, it helps improve the readability and
understanding of the production rules.

Modularity

The modularity of a production rule helps in its incremental

improvement as the production rule can be in discrete parts. The
production rule is made from a collection of information and facts
that may not have dependencies unless there is a rule connecting
them together. The addition or deletion of single information will not
have a major effect on the output. Modularity helps enhance the
performance of the production system by adjusting the parameters
of the rules.

Modifiability

The feature of modifiability helps alter the rules as per requirements. Initially, the skeletal form of the production system is created. We then gather
the requirements and make changes in the raw structure of the production system. This helps in the iterative improvement of the production system.

Knowledge-intensive

Production systems contain knowledge in the form of a human spoken language, i.e., English. It is not built using any programming languages. The
knowledge is represented in plain English sentences. Production rules help make productive conclusions from these sentences.
Negative Characteristics of a Production System in AI

Despite strengths, certain characteristics of a production system in AI – listed below – can

cause difficulties. Larger rule sets or overlapping conditions may turn a neat system into a
demanding one.

•Opacity: With dozens or even hundreds of rules firing over time, it can become tough to
trace which rule caused a particular result. That confusion may slow down debugging.

•Inefficiency: Some systems check all rules each time new facts appear. As the rule set
grows, so can the time it takes to find and apply the right actions.

•Lack of Autonomous Learning: Unless combined with learning tools, these systems don’t
create or modify rules on their own. They only do what the predefined statements allow.

•Conflict Resolution Overhead: When multiple rules match at once, the system must pick
one. Deciding how to break ties can add complexity and affect performance.
Intelligent agents in AI are autonomous entities that act upon an environment using sensors and actuators to achieve their
goals. In addition, intelligent agents may learn from the environment to achieve those goals. Driverless cars and the Siri
virtual assistant are examples of intelligent agents in AI.

These are the main four rules all AI agents must adhere to:

•Rule 1: An AI agent must be able to perceive the environment.

•Rule 2: The environmental observations must be used to make decisions.

•Rule 3: The decisions should result in action.

•Rule 4: The action taken by the AI agent must be a rational. Rational actions are actions that maximize performance and
yield the best positive outcome.
Autonomous Vehicles
Even though these cars
span multiple types of
intelligent agents, they’re
a good example of model-
based reflex agents.
Complex systems like
traffic and pedestrian
movements are exactly the
kind of challenge that
model-based reflex agents
are designed for.
Their internal model is
used to make real-time
decisions on the road, like
braking when another car
runs a red light, or slowing
down rapidly when the car
ahead does the same.
Their internal system is
constantly updating based
 A real-life example of a goal-based agent is
a GPS navigation system in a car; it
Key points about goal-based agents:
continuously monitors the current location
•Defined goal: and road conditions to choose the best
route to reach a predetermined destination
•They have a clear objective or desired outcome that they
strive to achieve. (the goal), adapting to changes in traffic or
road closures to achieve the desired
•Planning and decision-making: outcome.
•They evaluate potential actions and their consequences to Other examples include:
choose the most likely path to reach the goal. A personal fitness app:
This app would analyze your fitness goals
•Environmental awareness:
and create personalized workout plans
based on your current health data, aiming to
•They continuously monitor their environment to adapt their
actions as needed.
achieve your desired fitness level.
A robotic vacuum cleaner:
It navigates a room with the goal of cleaning
the floor, adjusting its movement based on
obstacles and dirt detection.
Types of search algorithms:
Depth First Search:

Depth-first search (DFS) is an algorithm for traversing or searching tree or graph data structures. The algorithm starts at the root node (selecting some
arbitrary node as the root node in the case of a graph) and explores as far as possible along each branch before backtracking. It uses last in- first-out
strategy and hence it is implemented using a stack.