Artificial Intelligence

▌Majid Ramezani
Institute for Advanced Studies in Basic Science (IASBS)
ramezani@[Link]; [Link]@[Link]
Source Materials
■ No textbook is required. Readings will come from freely available online materials.

■ Those who may be interested to read additional references, would refer to:

Stuart Russell, Peter Norvig, Artificial Intelligence: A Modern Approach, Pearson, 2009.
David L. Poole & Alan K. Mackworth, Artificial Intelligence: Foundations of Computational Agents, Cambridge
University Press, 2010.
Deepak Khemani, A First Course in Artificial Intelligence, Mc Graw Hill India, 2008.
Denis Rothman, Artificial Intelligence By Example-2nd edition, Packt Publishing, 2020.
Wolfgang Ertel, Introduction to Artificial Intelligence, Springer, 2017.
etc.
Evaluation
■ Endterm exam (30 %)
■ Midterm exam (20 %)
■ Homework (25 %)
• Both theory and programming
• Collaboration policy:
i. first try to solve the problems on your own,
ii. then, can discuss with other classmates,
iii. write-up solutions on your own.

■ Project (25 %)
• Logistics :
i. up to 2 students per group,
ii. the problem and corresponding deadline, will set after passing the midterm exam,
iii. in case of creative ideas, the problem will be entrusted to students.

Artificial Intelligence (Majid Ramezani)

█ Course Outline
Part I: Artificial Intelligence
1. Introduction
Part II: Problem-solving
2. Uninformed (Blind) Search; solving problems by searching
3. Informed (Heuristic) Search; search in complex environments
4. Metaheuristic Search; solving optimization problems with advanced strategies
5. Adversarial Search and Games
Part III: Knowledge, reasoning, and planning
6. Logical Agents
7. First-Order Logic
8. Inference in First-Order Logic
9. Knowledge Representation
Part IV: Uncertain knowledge and reasoning
10. Quantifying Uncertainty
11. Probabilistic Reasoning
12. Probabilistic Programming
Artificial Intelligence (Majid Ramezani)
█ Course Outline
Part I: Artificial Intelligence
1. Introduction
Part II: Problem-solving
2. Uninformed (Blind) Search; solving problems by searching
3. Informed (Heuristic) Search; search in complex environments
4. Metaheuristic Search; solving optimization problems with advanced strategies
5. Adversarial Search and Games
Part III: Knowledge, reasoning, and planning
6. Logical Agents
7. First-Order Logic
8. Inference in First-Order Logic
9. Knowledge Representation
Part IV: Uncertain knowledge and reasoning
10. Quantifying Uncertainty
11. Probabilistic Reasoning
12. Probabilistic Programming
Artificial Intelligence (Majid Ramezani)
ARTIFICIAL
INTELLIGENCE
Chapter 1:
Introduction
This chapter includes rudimentary and basic
concepts about artificial intelligence (AI).
At the end of this chapter, one should find that:

 What is the idea behind AI, and AI definition,

 The branches of AI and their goals,
 Agents in AI.
Chapter 1: Introduction
Lecture Outline
1-1 What is Artificial Intelligence (AI)? 1-4 Applications of AI
1-1-1 AI Definition 1-5 Agents
1-1-2 AI History 1-5-1 Single-Agent Systems vs. Multi-Agent Systems

1-2 Goals of AI 1-5-2 Examples of Agents

1-3 Branches of AI 1-5-3 Environment

1-3-1 Machine Learning (ML)

1-3-2 Natural Language Processing (NLP)

1-3-3 Digital Image Processing & Computer Vision

1-3-4 Robotics

1-3-5 Expert Systems

1-3-6 Evolutionary Algorithms (EA)
1-3-7 Fuzzy Logic
1-3-8 Other Areas

Artificial Intelligence (Chapter 1: Introduction)

Chapter 1: Introduction
Lecture Outline
1-1 What is Artificial Intelligence (AI)? 1-4 Applications of AI
1-1-1 AI Definition 1-5 Agents
1-1-2 AI History 1-5-1 Single-Agent Systems vs. Multi-Agent Systems

1-2 Goals of AI 1-5-2 Examples of Agents

1-3 Branches of AI 1-5-3 Environment

1-3-1 Machine Learning (ML)

1-3-2 Natural Language Processing (NLP)

1-3-3 Digital Image Processing & Computer Vision

1-3-4 Robotics

1-3-5 Expert Systems

1-3-6 Evolutionary Algorithms (EA)
1-3-7 Fuzzy Logic
1-3-8 Other Areas

Artificial Intelligence (Chapter 1: Introduction)

1-1 What is Artificial Intelligence (AI)?
Understanding the Real Meaning of AI
In today's world, the term “Artificial Intelligence” is often surrounded by hype, misconceptions, and exaggerated expectations.
Many people think of AI as something out of a science fiction movie—like robots taking over the world or machines that think and
feel like humans. These ideas, mostly shaped by popular media. However, these perceptions can distort our understanding of what
AI truly is and what it aims to achieve.

Artificial Intelligence, at its core, is not about creating machines that think or feel like humans. Instead, it is a branch of computer
science dedicated to building systems that can perform tasks requiring human intelligence—such as learning, reasoning, and
problem-solving—without necessarily mimicking the human mind itself.

Artificial Intelligence (Chapter 1: Introduction) 1 / 97

Chapter 1: Introduction
Lecture Outline
1-1 What is Artificial Intelligence (AI)? 1-4 Applications of AI
1-1-1 AI Definition 1-5 Agents
1-1-2 AI History 1-5-1 Single-Agent Systems vs. Multi-Agent Systems

1-2 Goals of AI 1-5-2 Examples of Agents

1-3 Branches of AI 1-5-3 Environment

1-3-1 Machine Learning (ML)
1-3-2 Natural Language Processing (NLP)

1-3-3 Digital Image Processing & Computer Vision

1-3-4 Robotics

1-3-5 Expert Systems

1-3-6 Evolutionary Algorithms (EA)
1-3-7 Fuzzy Logic
1-3-8 Other Areas

Artificial Intelligence (Chapter 1: Introduction)

1-1-1 AI Definition
Breaking down the term "Artificial Intelligence" into its constituent parts helps in understanding the foundational
concepts before combining them into a comprehensive definition;

■ Artificial, means what made or produced by human beings rather than occurring naturally (namely, Man-Made Origin).
Artificial light (e.g., electric lighting vs. sunlight)

■ Intelligence, is the ability to learn, solve problems, reason, understand, be self-aware, and apply knowledge and skills. It
encompasses capacities such as reasoning, problem-solving, perception, creativity, and learning from experience [1].

[1] Flynn, James R. What is intelligence?: Beyond the Flynn effect. Cambridge University Press, 2007.

Artificial Intelligence (Chapter 1: Introduction) 2 / 97

1-1-1 AI Definition (continued…)
Let’s define “Artificial Intelligence” according to the fathers and pioneers of AI and computer science.

▌Artificial Intelligence (AI) is the study of agents that receive percepts from the environment and perform actions [2].
Stuart Russell

▌The science and engineering of making intelligent machines, especially intelligent computer programs. It is related to
the similar task of using computers to understand human intelligence, but AI does not have to confine itself to methods
that are biologically observable [3].
John McCarthy

▌Artificial intelligence is the science of making machines do things that would require intelligence if done by men [4].
Marvin Minsky

[2] Russell, Stuart J., and Peter Norvig. Artificial intelligence: a modern approach. Pearson, 2016.
[3] McCarthy, John. "From here to human-level AI." Artificial Intelligence 171, no. 18 (2007): 1174-1182.
[4] Minsky, Marvin. "1968, Semantic information processing."

Artificial Intelligence (Chapter 1: Introduction) 3 / 97

1-1-1 AI Definition (continued…)
There is no accepted definition of artificial intelligence, but lets to have our definition:

▌Artificial intelligence (AI) is technology that enables computers and machines to simulate human learning, comprehension,
problem solving, decision making, creativity and autonomy.

Namely, AI is the field of computer science and engineering focused on creating machines and software capable of performing tasks
that typically require human intelligence. This involves designing systems that can perceive their environment, reason, learn, and
make decisions or take actions to achieve specific goals.

Artificial Intelligence (Chapter 1: Introduction) 4 / 97

Chapter 1: Introduction
Lecture Outline
1-1 What is Artificial Intelligence (AI)? 1-4 Applications of AI
1-1-1 AI Definition 1-5 Agents
1-1-2 AI History 1-5-1 Single-Agent Systems vs. Multi-Agent Systems

1-2 Goals of AI 1-5-2 Examples of Agents

1-3 Branches of AI 1-5-3 Environment

1-3-1 Machine Learning (ML)

1-3-2 Natural Language Processing (NLP)

1-3-3 Digital Image Processing & Computer Vision

1-3-4 Robotics

1-3-5 Expert Systems

1-3-6 Evolutionary Algorithms (EA)
1-3-7 Fuzzy Logic
1-3-8 Other Areas

Artificial Intelligence (Chapter 1: Introduction)

1-1-2 AI History
The Birth of AI (1950s)
▌Alan Turing and the Turing Test (1950): Alan Turing, often regarded as the father of AI, proposed the
idea of a “thinking machine” in his seminal paper “Computing Machinery and Intelligence”. He
introduced the Turing Test as a measure of a machine's ability to exhibit intelligent behavior
indistinguishable from that of a human.

Turing Test: If a machine could carry on a conversation (over a teleprinter)

that was indistinguishable from a conversation with a human being, then it
was reasonable to say that the machine was “thinking”. The Turing Test was
the first serious proposal in the philosophy of artificial intelligence.

Artificial Intelligence (Chapter 1: Introduction) 5 / 97

1-1-2 AI History (continued…)
The Dartmouth Conference (1956)
▌The official birth of AI as a field is often marked by the Dartmouth Conference, where John McCarthy, Marvin Minsky, Nathaniel
Rochester, and Claude Shannon coined the term “Artificial Intelligence” and discussed the possibility of creating intelligent
machines. This event set the stage for the next several decades of AI research.

Artificial Intelligence (Chapter 1: Introduction) 6 / 97

1-1-2 AI History (continued…)
Early AI Research and Optimism (1950s-1970s)
▌Symbolic AI and Early Programs: Early AI research focused on symbolic AI, also known as “Good Old-Fashioned AI” (GOFAI), which
involved manipulating symbols (rather than numerical data) to represent knowledge and solve problems. Notable early programs
included the Logic Theorist (1955), developed by Allen Newell and Herbert A. Simon, and the General Problem Solver (1957).
▌Rise of Expert Systems: In the late 1960s and 1970s, AI research saw the development of expert systems, which used rule-based
systems to emulate the decision-making abilities of a human expert in specific domains (e.g., MYCIN for medical diagnosis).

The AI Winters (1970s-1990s)

▌First AI Winter (1974-1980): Due to unmet expectations and limited computational power, funding for AI research diminished
significantly, leading to a period known as the “AI Winter”.
▌Second AI Winter (1987-1993): Despite some advances, the collapse of the Lisp machine market (hardware specifically designed for
AI) and growing skepticism led to another downturn in AI research and investment.

Artificial Intelligence (Chapter 1: Introduction) 7 / 97

1-1-2 AI History (continued…)
Revival and the Rise of Machine Learning (1990s-2000s)
▌Resurgence in AI Research: the late 1980s and 1990s saw a revival of interest in AI, fueled by advances in computational power, the
development of backpropagation for training neural networks, and the emergence of new algorithms in machine learning.

▌The Success of Deep Blue (1997): IBM's Deep Blue made history by defeating world chess champion Garry Kasparov, showcasing the
potential of AI in specialized tasks.

▌The Rise of Data-Driven Approaches: with the explosion of data and the advent of the internet, AI research shifted focus towards
data-driven methods, particularly in machine learning and statistical approaches, marking a move away from symbolic AI.

The Age of Deep Learning and AI in the Real World (2010s-Present)

▌Deep Learning Breakthroughs: The 2010s witnessed the rise of deep learning, a subset of machine learning that utilizes multi-
layered neural networks to perform complex tasks such as image and speech recognition. Landmark achievements include Google's
AlphaGo defeating world champion Go player Lee Sedol in 2016.
▌AI in Everyday Life: AI has become integrated into many aspects of daily life, from virtual assistants like Siri and Alexa to
recommendation systems on platforms like Netflix and Amazon.
▌Generative AI: the late 2010s and early 2020s have seen the rise of generative AI models like GPT (Generative Pre-trained
Transformer) and DALL-E, which can generate human-like text and images.

Artificial Intelligence (Chapter 1: Introduction) 8 / 97

Chapter 1: Introduction
Lecture Outline
1-1 What is Artificial Intelligence (AI)? 1-4 Applications of AI
1-1-1 AI Definition 1-5 Agents
1-1-2 AI History 1-5-1 Single-Agent Systems vs. Multi-Agent Systems

1-2 Goals of AI 1-5-2 Examples of Agents

1-3 Branches of AI 1-5-3 Environment

1-3-1 Machine Learning (ML)

1-3-2 Natural Language Processing (NLP)

1-3-3 Digital Image Processing & Computer Vision

1-3-4 Robotics

1-3-5 Expert Systems

1-3-6 Evolutionary Algorithms (EA)
1-3-7 Fuzzy Logic
1-3-8 Other Areas

Artificial Intelligence (Chapter 1: Introduction)

1-2 Goals of AI
The primary goal of AI is to create systems that can perform tasks requiring human intelligence. This includes a wide range of activities
such as understanding natural language, recognizing patterns, making decisions, solving problems, and learning from experience.

▌Key Aspects of the Primary Goal:

■ Automation of Intellectual Tasks: the core aim is to develop AI systems that can automate tasks that typically require human
cognitive abilities. This includes everything from recognizing images and understanding speech to making decisions and translating
languages.

■ Enhancing and Extending Human Capabilities: AI seeks to augment human intelligence by providing tools that help people solve
complex problems, make better decisions, and enhance productivity in various domains, such as medicine, finance, and education.

■ Understanding and Replicating Cognitive Processes: a major focus of AI is to understand the underlying principles of human
cognition and replicate those processes in machines. This involves creating algorithms that can mimic learning, reasoning, and
problem-solving abilities.

■ Adaptability and Learning: AI systems are designed to learn from data and experience, improving
their performance over time without needing explicit programming for each new task. This
adaptability is key to making AI useful in dynamic and complex environments.

Artificial Intelligence (Chapter 1: Introduction) 9 / 97

1-2 Goals of AI (continued…)
Be careful that not everything related to design and development will result in intelligent agents.

(-Ꝏ , + Ꝏ)

......

......

Artificial Intelligence (Chapter 1: Introduction) 10 / 97

Chapter 1: Introduction
Lecture Outline
1-1 What is Artificial Intelligence (AI)? 1-4 Applications of AI
1-1-1 AI Definition 1-5 Agents
1-1-2 AI History 1-5-1 Single-Agent Systems vs. Multi-Agent Systems

1-2 Goals of AI 1-5-2 Examples of Agents

1-3 Branches of AI 1-5-3 Environment

1-3-1 Machine Learning (ML)

1-3-2 Natural Language Processing (NLP)

1-3-3 Digital Image Processing & Computer Vision

1-3-4 Robotics

1-3-5 Expert Systems

1-3-6 Evolutionary Algorithms (EA)
1-3-7 Fuzzy Logic
1-3-8 Other Areas

Artificial Intelligence (Chapter 1: Introduction)

1-3 Branches of AI
To achieve its overarching goal of replicating and extending human intelligence, AI has been divided into several specialized branches,
each tackling specific aspects of the broader challenge. These branches focus on different sub-problems, such as learning from data,
understanding natural language, perceiving the environment, and making decisions.
By breaking down the general problem of creating intelligent systems into smaller, more manageable problems, AI research can address
the complexities of each area more effectively. Each branch, from machine learning to robotics, plays a crucial role in advancing the
capabilities of AI, ultimately contributing to the realization of its primary goals.

The most critical branches of AI, include:

● Machine Learning (ML)

● Natural Language Processing (NLP)

● Digital Image Processing & Computer Vision

● Robotics

● Expert Systems

● Optimization

● Fuzzy Logic

Artificial Intelligence (Chapter 1: Introduction) 11 / 97

Chapter 1: Introduction
Lecture Outline
1-1 What is Artificial Intelligence (AI)? 1-4 Applications of AI
1-1-1 AI Definition 1-5 Agents
1-1-2 AI History 1-5-1 Single-Agent Systems vs. Multi-Agent Systems

1-2 Goals of AI 1-5-2 Examples of Agents

1-3 Branches of AI 1-5-3 Environment

1-3-1 Machine Learning (ML)

1-3-2 Natural Language Processing (NLP)

1-3-3 Digital Image Processing & Computer Vision

1-3-4 Robotics

1-3-5 Expert Systems

1-3-6 Evolutionary Algorithms (EA)
1-3-7 Fuzzy Logic
1-3-8 Other Areas

Artificial Intelligence (Chapter 1: Introduction)

1-3-1 Machine Learning (ML)
Machine learning is a branch of AI that aims to replicate human learning processes in machines. It enables systems to learn from data,
identify patterns, and make decisions with minimal human intervention, striving to perform tasks in a manner that closely resembles
human cognitive functions.

So lets take a look to “human learning”.

One method of human learning is concept learning, which involves understanding and categorizing information based on observations
and experiences. This approach enables individuals to form mental representations of categories and generalize from specific instances
to broader concepts.

Artificial Intelligence (Chapter 1: Introduction) 12 / 97

1-3-1 Machine Learning (ML) (continued…)
Concept learning encompasses learning how to discriminate, recognize, and categorize things (with critical
attributes and features) [5-7]

How many types of trees are there in the world?!

[5] Hunt, Earl B. "The nature of concept learning." (1962).

[6] Lake, Brenden M., Ruslan Salakhutdinov, and Joshua B. Tenenbaum. "Human-level concept learning through probabilistic program induction." Science 350, no. 6266 (2015): 1332-1338.
[7] Zeithamova, Dagmar, Michael L. Mack, Kurt Braunlich, Tyler Davis, Carol A. Seger, Marlieke TR Van Kesteren, and Andreas Wutz. "Brain mechanisms of concept learning." Journal of Neuroscience 39, no. 42 (2019): 8259-8266.

Artificial Intelligence (Chapter 1: Introduction) 13 / 97

1-3-1 Machine Learning (ML) (continued…)

How many types of cars are there in the world?!

Artificial Intelligence (Chapter 1: Introduction) 14 / 97

1-3-1 Machine Learning (ML) (continued…)

How many types of flowers are there in the world?!

It is not possible to provide an exact definition for a flower or tree, or any similar concept!

Artificial Intelligence (Chapter 1: Introduction) 15 / 97

1-3-1 Machine Learning (ML) (continued…)
Generally, there are miscellaneous samples in the world that it is not possible to provide a fixed and exact definition for
them. In such cases, concept learning help human to recognize, distinguish, and categorize them throughout a set of
features and attributes.

In fact, “they are too complicated to be defined exactly” !!

Human daily decision makings, are severely involved with concept learning.

Artificial Intelligence (Chapter 1: Introduction) 16 / 97

1-3-1 Machine Learning (ML) (continued…)
▌Natural Neural Networks

What is the brain? The brain is a complex organ that controls thought, memory,
emotion, touch, motor skills, vision, breathing, temperature, hunger and every
process that regulates our body [8]

Neurons are the key players in the brain. They are the fundamental units of the
brain and nervous system, the cells responsible for receiving sensory input from the
external world, for sending motor commands to our muscles, and for transforming
and relaying the electrical signals at every step in between.

[8] Brain Anatomy and How the Brain Works, Johns Hopkins Medicine, 2023.

Artificial Intelligence (Chapter 1: Introduction) 17 / 97

1-3-1 Machine Learning (ML) (continued…)
■ Human brain roughly consists of 𝟏𝟎𝟏𝟏 neurons, and 𝟏𝟎𝟏𝟔 connections among them.

■ There are three types of neurons in human body:

• Sensory neurons: sensory neurons are the nerve cells that are activated by sensory input from the environment - for example, when you touch a
hot surface with your fingertips, the sensory neurons will be the ones firing and sending off signals to the rest of the nervous system about the
information they have received.

• Motor neurons: motor neurons of the spinal cord are part of the central nervous system (CNS) and connect to muscles, glands and organs
throughout the body. These neurons transmit impulses from the spinal cord to skeletal and smooth muscles (such as those in your stomach), and so
directly control all of our muscle movements. There are in fact two types of motor neurons: those that travel from spinal cord to muscle are called
lower motor neurons, whereas those that travel between the brain and spinal cord are called upper motor neurons.

• Interneurons: as the name suggests, interneurons are the ones in between - they connect spinal motor and sensory neurons. As well as transferring
signals between sensory and motor neurons, interneurons can also communicate with each other, forming circuits of various complexity. They are
multipolar, just like motor neurons.

Artificial Intelligence (Chapter 1: Introduction) 18 / 97

1-3-1 Machine Learning (ML) (continued…)
The structure of a neuron has three main parts:

■ Cell body (or soma): the cell body is the core section of the neuron. The
cell body contains genetic information, maintains the neuron’s structure,
and provides energy to drive activities. It’s enclosed by a membrane that
both protects it and allows it to interact with its immediate
surroundings.

■ Axon: the transmitting part of a neuron. In practice, it carries nerve

impulses away from the cell body.

■ Dendrite: are fibrous roots that branch out from the cell body. Like
antennae, dendrites receive and process signals from the axons of other
neurons. Neurons can have more than one set of dendrites, known as
dendritic trees. In practice, they receive the electrical messages (from
other neurons) and transfer them to the soma.

A synapse is a small gap between two neurons, where nerve impulses are relayed by a neurotransmitter from the
axon of a presynaptic (sending) neuron to the dendrite of a postsynaptic (receiving) neuron.

Artificial Intelligence (Chapter 1: Introduction) 19 / 97

1-3-1 Machine Learning (ML) (continued…)

A sequence of two neurons:

Artificial Intelligence (Chapter 1: Introduction) 20 / 97

1-3-1 Machine Learning (ML) (continued…)
The central nervous system (CNS which is made up of the brain and spinal cord) is the body's processing center. It spread
throughout the human body.

Artificial Intelligence (Chapter 1: Introduction) 21 / 97

1-3-1 Machine Learning (ML) (continued…)
An animation of signal transmission among human nervous system.

Artificial Intelligence (Chapter 1: Introduction) 22 / 97

1-3-1 Machine Learning (ML) (continued…)
▌The function of a neuron
A given neuron receives hundreds of inputs, almost exclusively on its dendrites and cell body. These inputs add and subtract in
a constantly evolving pattern, depending on what the brain is thinking. This is a process called synaptic integration, which
determines whether a neuron becomes active.
In order to become active, the total input must reach a threshold at which excitation outweighs inhibition enough. Only at this
point will the receiving neuron spike, adding its voice to the conversation by releasing its own neurotransmitter.

A neuron spikes when a combination of all the excitation and inhibition it receives makes it reach threshold. On the right is an example from an
actual neuron in the mouse's cortex. (Image: Alan Woodruff / QBI)

Artificial Intelligence (Chapter 1: Introduction) 23 / 97

1-3-1 Machine Learning (ML) (continued…)
A given neuron receives hundreds of inputs, almost exclusively on its dendrites and cell body. These inputs add and subtract in
a constantly evolving pattern, depending on what the brain is thinking. This is a process called synaptic integration, which
determines whether a neuron becomes active.

Artificial Intelligence (Chapter 1: Introduction) 24 / 97

1-3-1 Machine Learning (ML) (continued…)
Take a look inside your mind!

Artificial Intelligence (Chapter 1: Introduction) 25 / 97

1-3-1 Machine Learning (ML) (continued…)
A sample animation of neuron firing in human mind and body:

Artificial Intelligence (Chapter 1: Introduction) 26 / 97

1-3-1 Machine Learning (ML) (continued…)
▌A better definition for “learning” in human

Each and every time we learn something new, our brain forms new connections and neurons and makes existing neural
pathways stronger or weaker. Some experts call these changes “plasticity” in the brain.

Therefore, it can be stated that learning would be paraphrased as the formation of new string of neurons in human mind.

☺ Think about Alzheimer…..

Artificial Intelligence (Chapter 1: Introduction) 27 / 97

1-3-1 Machine Learning (ML) (continued…)
▌Leaning in “machine”

Machine Learning (ML) is a branch of artificial intelligence and computer science which focuses on the use of
data and algorithms to imitate the way that humans learn, gradually improving its accuracy.

Artificial Intelligence (Chapter 1: Introduction) 28 / 97

1-3-1 Machine Learning (ML) (continued…)
One can break out the learning system of a machine learning algorithm into three main parts:

Arthur Samuel (1959): Machine Learning is the field of study that gives the computer the
ability to learn without being explicitly programmed.

Tom Mitchell (1998): a computer program is said to learn from experience E with respect to
some class of tasks T and performance measure P, if its performance at tasks in T, as measured
by P, improves with experience E.

Artificial Intelligence (Chapter 1: Introduction) 29 / 97

1-3-1 Machine Learning (ML) (continued…)
One can breaks out the learning system of a machine learning algorithm into three main parts.

■ A Decision Process: In general, machine learning algorithms are used to make a prediction or classification. Based on some input data,
which can be labeled or unlabeled, your algorithm will produce an estimate about a pattern in the data.

■ An Error Function: An error function evaluates the prediction of the model. If there are known examples, an error function can make a
comparison to assess the accuracy of the model.

■ A Model Optimization Process: If the model can fit better to the data points in the training set, then weights are adjusted to reduce the
discrepancy between the known example and the model estimate. The algorithm will repeat this “evaluate and optimize” process,
updating weights autonomously until a threshold of accuracy has been met.

Artificial Intelligence (Chapter 1: Introduction) 30 / 97

1-3-1 Machine Learning (ML) (continued…)

The general workflow in machine learning algorithms would be as follows:

Artificial Intelligence (Chapter 1: Introduction) 31 / 97

1-3-1 Machine Learning (ML) (continued…)
The exact workflow is presented here. In future (next semester), you will understand this chart better.

Artificial Intelligence (Chapter 1: Introduction) 32 / 97

1-3-1 Machine Learning (ML) (continued…)
▌Concept Learning
Like human, the machine also utilize the “concept learning”. As it described, it is so useful in complicated problems.

As a kind of observational learning, in concept learning:

■ Data is a crucial component in concept learning (and hence, in the field of ML). It refers to the set of observations or
measurements that can be used to train a machine-learning model.

■ The quality and quantity of data available for training and testing play a significant role in determining the
performance of a machine-learning model.

Artificial Intelligence (Chapter 1: Introduction) 33 / 97

1-3-1 Machine Learning (ML) (continued…)
As different tasks would be tackled by ML, therefore a miscellaneous types of data would be applicable, such as:

■ Images in different image (or video) processing.

■ Text in a variety of natural language processing tasks.

■ Facts in different logic or knowledge driven tasks.

■ Numeric values

■ etc.

Artificial Intelligence (Chapter 1: Introduction) 34 / 97

1-3-1 Machine Learning (ML) (continued…)
Like human, the machine also utilize the “concept learning”. As it described, it is so useful in complicated problems.

■ Human daily decision makings, are severely involved with concept learning.

■ Likewise, several successful and even astonishing tasks have been performed by machine through concept learning
methods, such as:

Artificial Intelligence (Chapter 1: Introduction) 35 / 97

1-3-1 Machine Learning (ML) (continued…)
Face recognition: a sample of concept learning application in machine.

After training the model…

Some training sample from an cat-dog classification dataset.

Cat-Dog Classification Model

Artificial Intelligence (Chapter 1: Introduction) 36 / 97

1-3-1 Machine Learning (ML) (continued…)
Face recognition: a sample of concept learning application in machine.

After training the model…

Some training sample from an image recognition dataset.

Face Recognition

Artificial Intelligence (Chapter 1: Introduction) 37 / 97

1-3-1 Machine Learning (ML) (continued…)
A sample of concept learning application in machine.

After training the model…

Weather Prediction

Artificial Intelligence (Chapter 1: Introduction) 38 / 97

1-3-1 Machine Learning (ML) (continued…)
Another sample of concept learning application in machine.

Artificial Intelligence (Chapter 1: Introduction) 39 / 97

1-3-1 Machine Learning (ML) (continued…)
Another sample of concept learning application in machine.

Artificial Intelligence (Chapter 1: Introduction) 40 / 97

1-3-1 Machine Learning (ML) (continued…)
A sample of concept learning application in machine.

Artificial Intelligence (Chapter 1: Introduction) 41 / 97

1-3-1 Machine Learning (ML) (continued…)
Another sample of concept learning application in machine.

Artificial Intelligence (Chapter 1: Introduction) 42 / 97

1-3-1 Machine Learning (ML) (continued…)
▌Artificial Neural Networks (ANN)
In fact an Artificial Neural Network (ANN) is a network of simulated natural neurons. Let to consider the function of a single
artificial neuron.

Artificial Intelligence (Chapter 1: Introduction) 43 / 97

1-3-1 Machine Learning (ML) (continued…)
What happens in a single artificial neuron:

A single artificial neuron

Artificial Intelligence (Chapter 1: Introduction) 44 / 97

1-3-1 Machine Learning (ML) (continued…)
Like neurons in human nervous system, each artificial neuron is connected to other neurons via a biological
axon-synapse-dendrite connection.

A sample artificial neural network

Artificial Intelligence (Chapter 1: Introduction) 45 / 97

1-3-1 Machine Learning (ML) (continued…)
The weights of the connections in an ANN model, will fix after several iterations. Then, the model would be used for user
inputs.

A sample ANN model

Artificial Intelligence (Chapter 1: Introduction) 46 / 97

1-3-1 Machine Learning (ML) (continued…)
For example in a numeric character recognition (from image) task:

Numeric character recognition (from image), sample input image for “7”

It can be claimed that, learning an ANN model is just setting the weights of the connections in the model,
appropriately.

Artificial Intelligence (Chapter 1: Introduction) 47 / 97

1-3-1 Machine Learning (ML) (continued…)
▌Types of learning in machine
Considering the learning styles in machine learning algorithms, there are three types of algorithms:
1. Supervised Learning: in which the training data is labeled, such as spam/not-spam or cat/dog image classification, etc.
A model is prepared through a training process in which it is required to make predictions and is corrected when those predictions are wrong.
The training process continues until the model achieves a desired level of accuracy on the training data.

2. Unsupervised Learning: in which the training data is not labeled.

A model is prepared by deducing structures present in the input data. This may be to extract general rules or hidden patterns. It may be through
a mathematical process to systematically reduce redundancy, or it may be to organize data by similarity.

DOG

DOG

CAT

Artificial Intelligence (Chapter 1: Introduction) 48 / 97

1-3-1 Machine Learning (ML) (continued…)

Artificial Intelligence (Chapter 1: Introduction) 49 / 97

1-3-1 Machine Learning (ML) (continued…)
3. Semi-supervised Learning: in which the dataset is a mixture of labeled and un-labelled samples.
Semi-supervised learning bridges supervised learning and unsupervised learning techniques to solve their key challenges. With it, you train an
initial model on a few labeled samples and then iteratively apply it to the greater number of unlabeled data

■ Unlike unsupervised learning, SSL works for a variety of problems from classification and regression to clustering and
association.
■ Unlike supervised learning, the method uses small amounts of labeled data and also large amounts of unlabeled data, which
reduces expenses on manual annotation and cuts data preparation time.

ML Model Prediction

Since unlabeled data is abundant, easy to get, and cheap, semi-supervised learning finds many applications, while the accuracy of results doesn’t
suffer.
[Link]

Artificial Intelligence (Chapter 1: Introduction) 50 / 97

1-3-1 Machine Learning (ML) (continued…)
Supervised learning can be categorized into two problems, depending on the nature of the target variable
(label). These include classification and regression.
■ Classification: focuses on sorting observations into distinct categories. It is when the target label is a category, namely it has a discrete
value. Such as “blue” or “red” or “disease” and “no disease.” An example of a classification problem is a spam filter that classifies emails
into two categories, ‘safe’ or ‘suspicious.’

■ Regression: is based on making predictions of continuous target variables. Examples of regression problems include using historical
stock market returns to forecast stock price performance or predicting the number of goals scored by a team.

Artificial Intelligence (Chapter 1: Introduction) 51 / 97

1-3-1 Machine Learning (ML) (continued…)
Unsupervised learning also can be categorized into two problems. These include clustering and association rules.

■ Clustering: It mainly deals with finding a structure or pattern in a

collection of uncategorized data. K-means, is an example of
clustering method.

■ Association Rules: An association rule is a rule-based method for

finding relationships between variables in a given dataset. These
methods are frequently used for market basket analysis, allowing
companies to better understand relationships between different
products. Apriori algorithm is an example of such methods.

Artificial Intelligence (Chapter 1: Introduction) 52 / 97

Chapter 1: Introduction
Lecture Outline
1-1 What is Artificial Intelligence (AI)? 1-4 Applications of AI
1-1-1 AI Definition 1-5 Agents
1-1-2 AI History 1-5-1 Single-Agent Systems vs. Multi-Agent Systems

1-2 Goals of AI 1-5-2 Examples of Agents

1-3 Branches of AI 1-5-3 Environment

1-3-1 Machine Learning (ML)

1-3-2 Natural Language Processing (NLP)

1-3-3 Digital Image Processing & Computer Vision

1-3-4 Robotics

1-3-5 Expert Systems

1-3-6 Evolutionary Algorithms (EA)
1-3-7 Fuzzy Logic
1-3-8 Other Areas

Artificial Intelligence (Chapter 1: Introduction)

1-3-2 Natural Language Processing (NLP)
▌What is Natural Language Processing (NLP)?
Natural Language Processing (NLP) is a field of artificial intelligence (AI) focused on enabling machines to understand, interpret, and
generate human language. It sits at the intersection of computer science, linguistics, and AI, and it encompasses a wide range of
tasks and techniques designed to bridge the gap between human communication and computer understanding. In fact, it aims to
enable machines to use language in the same way humans do.

It’s important to note that the term “Natural Language Processing” actually encompasses three distinct areas: natural
language understanding (NLU), natural language processing, and natural language generation (NLG). In fact, the term "NLP" is
somewhat of a misnomer.

Artificial Intelligence (Chapter 1: Introduction) 53 / 97

1-3-2 Natural Language Processing (NLP) (continued…)
▌Importance of NLP
● Remember AI’s ultimate goal!!
● Language is critical for human (as it forms his communication, thinking, etc.)
● NLP is crucial in the age of information, where massive amounts of text data are generated every day. It helps in automating
tasks, gaining insights from unstructured data, improving human-computer interaction, and making information more accessible
across different languages and formats.

Artificial Intelligence (Chapter 1: Introduction) 54 / 97

1-3-2 Natural Language Processing (NLP) (continued…)
▌Key Aspects of NLP:

● Text Processing and Understanding: text processing is a foundational aspect of NLP that involves the transformation and analysis
of written language to make it understandable and useful for machines. It encompasses several crucial phases: text pre-
processing (including, tokenization, stemming, lemmatization, etc.), and text processing (that meets the main task).

● Speech Processing: speech processing extends NLP into the realm of

spoken language, allowing machines to handle and interpret human
speech. This involves two main tasks: speech recognition (or Automatic
Speech Recognition - ASR that converts spoken language into text by
analyzing the acoustic signal and identifying the corresponding words), and
speech synthesis (text to Speech or TTS that generates spoken language
from written text. This involves modeling prosody -intonation, stress, and
rhythm- of speech to create natural-sounding speech).

● Language Models: Language models are statistical (traditional models like N-grams) and deep learning-based (like ChatGPT)
models that predict the likelihood of a sequence of words. They are crucial for many NLP tasks.

Artificial Intelligence (Chapter 1: Introduction) 55 / 97

1-3-2 Natural Language Processing (NLP) (continued…)
▌Main fields of NLP:

■ Linguistic Tasks: tries to tackle linguistics* concerns of language such as:

● Part-of-Speech Tagging (POS Tagging): labeling words in a sentence with their parts of speech (e.g., noun, verb).
● Syntax and Parsing: analyzing sentence structure to understand grammatical relationships.
● Stemming: reducing words to their root by removing prefixes or suffixes (e.g., "running" to "run").
● Morphological Analysis: understanding the structure of words and how they are formed.

* the scientific study of language and its structure, including the study of grammar, syntax, and phonetics. Specific branches of linguistics include sociolinguistics, dialectology,
psycholinguistics, computational linguistics, comparative linguistics, and structural linguistics.

Artificial Intelligence (Chapter 1: Introduction) 56 / 97

1-3-2 Natural Language Processing (NLP) (continued…)
▌Main fields of NLP: (continued….)

■ Text Processing: tries to tackle those tasks that are focused on textual inpts.

● Text Classification: categorizing text into predefined labels (e.g., spam detection, sentiment analysis).
● Named Entity Recognition (NER): identifying and classifying entities within text (e.g., names, dates, locations).
● Text Summarization: condensing large texts into shorter summaries while preserving key information.
● Machine Translation: automatically translating text from one language to another.
● Sentiment Analysis: determining the sentiment or emotional tone of a text.
● Text Generation: Producing coherent text based on input data or prompts (e.g., chatbots, creative writing).
● Topic Modeling: Discovering abstract topics that occur in a collection of documents.
● etc.

Artificial Intelligence (Chapter 1: Introduction) 57 / 97

1-3-2 Natural Language Processing (NLP) (continued…)
▌Main fields of NLP: (continued….)

■ Speech Processing: tries to tackle those tasks that are focused on audio speech inputs.

● Speech Recognition (Automatic Speech Recognition - ASR): converting spoken language into text.
● Speech Synthesis (Text-to-Speech - TTS): generating spoken language from written text.
● Speaker Identification and Verification: identifying or verifying a speaker based on their voice.
● Speech-to-Speech Translation: translating spoken language into another language in real-time.
● Voice Activity Detection: determining when speech occurs in an audio stream.
● Speech Summarization.

Artificial Intelligence (Chapter 1: Introduction) 58 / 97

Chapter 1: Introduction
Lecture Outline
1-1 What is Artificial Intelligence (AI)? 1-4 Applications of AI
1-1-1 AI Definition 1-5 Agents
1-1-2 AI History 1-5-1 Single-Agent Systems vs. Multi-Agent Systems

1-2 Goals of AI 1-5-2 Examples of Agents

1-3 Branches of AI 1-5-3 Environment

1-3-1 Machine Learning (ML)

1-3-2 Natural Language Processing (NLP)

1-3-3 Digital Image Processing & Computer Vision

1-3-4 Robotics

1-3-5 Expert Systems

1-3-6 Evolutionary Algorithms (EA)
1-3-7 Fuzzy Logic
1-3-8 Other Areas

Artificial Intelligence (Chapter 1: Introduction)

1-3-3 Digital Image Processing & Computer Vision
▌Digital Image Processing (DIP)

Digital Image Processing (DIP) involves the manipulation of digital images using computer algorithms. It focuses on the processing and
transformation of images to enhance their quality, extract useful information, or prepare them for further analysis. In fact, it aims to
enable machines to achieve human-like goals through visual perception.

Artificial Intelligence (Chapter 1: Introduction) 59 / 97

1-3-3 Digital Image Processing & Computer Vision (continued…)
▌What is the Importance of Digital Image Processing?

● Remember AI’s ultimate goal!!

● Enhancement and Visualization: enhances image quality for better visualization, which is critical in medical imaging, satellite
imagery, and photography.
● Information Extraction: enables the extraction of meaningful information from images, which is crucial for applications in remote
sensing, object recognition, and diagnostics.
● Preparation for Analysis: prepares images for further analysis in fields such as machine vision and computer vision, facilitating
tasks like feature extraction and object detection.
● Automation and Efficiency: automates tasks like image recognition and classification, improving efficiency and accuracy in various
industrial and research applications.

Artificial Intelligence (Chapter 1: Introduction) 60 / 97

1-3-3 Digital Image Processing & Computer Vision (continued…)
▌Subfields of Digital Image Processing

■ Image Enhancement: focuses on improving the visual appearance of an image or highlighting specific features. For example
contrast adjustment, histogram equalization, noise reduction.
■ Image Restoration: aims to recover the original image from a degraded version by removing noise, blurring, or other distortions.
For example de-blurring, noise filtering.
■ Image Segmentation: involves dividing an image into its constituent regions or objects for easier analysis. For example
thresholding, edge detection, region-based segmentation.
■ Image Compression: reduces the storage size of images without significant loss of quality, essential for efficient storage and
transmission. For examples JPEG, PNG.
■ Image Captioning: the primary goal of image captioning is to create a meaningful and accurate description of an image in natural
language. This can be used for various applications, including accessibility, content indexing, and enhancing human-computer
interaction.
■ Object Detection and Recognition: identifying and classifying objects within an image or video frame. For examples face
recognition, vehicle detection, landmark identification.
■ Image Classification: assigning a label or category to an entire image based on its content. For example classifying medical
images as healthy or diseased, identifying species in wildlife photography.

Artificial Intelligence (Chapter 1: Introduction) 61 / 97

1-3-3 Digital Image Processing & Computer Vision (continued…)
▌Subfields of Digital Image Processing

■ Image Segmentation: dividing an image into parts that correspond to different objects or regions, often a precursor to more
detailed analysis. For example semantic segmentation (labeling each pixel), instance segmentation (separating objects).
■ Pattern Recognition: detecting patterns and regularities in data, often used for recognizing specific features or textures in images.
For example handwriting recognition, texture classification.
■ Scene Understanding: interpreting a complex scene by recognizing the relationships between objects and their environment. For
example scene classification, context-aware recognition, holistic scene analysis.
■ Feature Extraction: identifying key points or regions in an image that are
informative and relevant for further analysis. For example SIFT, SURF, Harris
corners, edge detection.
■ etc.

Artificial Intelligence (Chapter 1: Introduction) 62 / 97

1-3-3 Digital Image Processing & Computer Vision (continued…)
▌Machine Vision (Computer Vision)

Encompasses a broader scope, including not just the analysis of individual images but also sequences of images or video frames. It
involves interpreting visual data over time and often includes understanding motion, tracking objects, and recognizing patterns across
frames.

Artificial Intelligence (Chapter 1: Introduction) 63 / 97

1-3-3 Digital Image Processing & Computer Vision (continued…)
▌Subfields of Machine Vision (Computer Vision)

■ Object Tracking: following the movement of objects across multiple frames or video sequences.
■ Motion Analysis: analyzing how objects or scenes change over time, including detecting and understanding movement.
■ Activity Recognition: identifying actions or events based on sequences of frames (e.g., detecting gestures, recognizing activities).
■ Video Surveillance: monitoring and analyzing video feeds for security and surveillance purposes.
■ etc.

Artificial Intelligence (Chapter 1: Introduction) 64 / 97

Chapter 1: Introduction
Lecture Outline
1-1 What is Artificial Intelligence (AI)? 1-4 Applications of AI
1-1-1 AI Definition 1-5 Agents
1-1-2 AI History 1-5-1 Single-Agent Systems vs. Multi-Agent Systems

1-2 Goals of AI 1-5-2 Examples of Agents

1-3 Branches of AI 1-5-3 Environment

1-3-1 Machine Learning (ML)

1-3-2 Natural Language Processing (NLP)

1-3-3 Digital Image Processing & Computer Vision

1-3-4 Robotics

1-3-5 Expert Systems

1-3-6 Evolutionary Algorithms (EA)
1-3-7 Fuzzy Logic
1-3-8 Other Areas

Artificial Intelligence (Chapter 1: Introduction)

1-3-4 Robotics
Robotics is an interdisciplinary field that integrates engineering, computer science, and various other disciplines to design, build, and
operate robots. Robots are machines capable of carrying out a series of actions automatically or semi-automatically.

Components of a Robot
 Sensors: devices that collect data from the environment (e.g., cameras, LIDAR, ultrasonic sensors).
 Actuators: mechanisms that perform actions based on sensor input (e.g., motors, servos).
 Controllers: The “brain” of the robot, typically a computer or microcontroller, that processes sensor data and sends commands to
actuators.
 Software: algorithms and programs that control the robot's behavior and decision-making.

Artificial Intelligence (Chapter 1: Introduction) 65 / 97

1-3-4 Robotics (continued…)
Types of Robots:
 Industrial Robots: Used in manufacturing for tasks such as assembly, welding, and painting.
 Service Robots: Designed for tasks in home or service environments (e.g., vacuum cleaners, medical robots).
 Autonomous Vehicles: Robots that navigate and perform tasks independently, such as drones and self-driving cars.

Artificial Intelligence (Chapter 1: Introduction) 66 / 97

1-3-4 Robotics (continued…)
▌What is the Importance of Robotics?

● Remember AI’s ultimate goal!!

● Automation and Efficiency: robots increase productivity by automating repetitive or hazardous tasks, leading to higher efficiency
and reduced human error in various industries.
● Precision and Consistency: robots perform tasks with high precision and consistency, which is crucial in fields such as
manufacturing, surgery, and quality control.
● Safety: robots can operate in dangerous environments, such as toxic chemical plants, deep-sea exploration, and outer space,
reducing risks to human operators.
● Assistance and Accessibility: Robots assist individuals with disabilities through assistive devices and enhance quality of life with
applications like robotic exoskeletons and prosthetics.
● Exploration and Innovation: robotics enables exploration of uncharted territories, including
space, deep oceans, and extreme environments, advancing scientific knowledge and
technological innovation.

Artificial Intelligence (Chapter 1: Introduction) 67 / 97

1-3-4 Robotics (continued…)
▌Subfields of Robotics

■ Robot Kinematics and Dynamics: studies the motion of robots and the forces involved. Kinematics deals with motion without
considering forces, while dynamics incorporates forces and torques.
■ Control Systems: focuses on designing algorithms and systems to control robot movements and behavior. This includes both low-
level control (e.g., motor control) and high-level control (e.g., path planning).
■ Robot Perception: involves the use of sensors and computer vision to enable robots to perceive and interpret their environment.
■ Robot Manipulation: deals with the physical interaction of robots with their environment, including tasks like grasping, handling,
and assembly.
■ Robotic Navigation and Localization: focuses on enabling robots to navigate through and understand their environment,
including mapping, localization, and path planning.
■ Human-Robot Interaction (HRI): Studies how robots interact with humans, including the design of interfaces and communication
methods.
■ Robotics in Healthcare: application of robotics in medical fields to assist with surgeries, rehabilitation, and patient care.
■ etc.

Artificial Intelligence (Chapter 1: Introduction) 68 / 97

1-3-4 Robotics (continued…)
▌Some samples….

Artificial Intelligence (Chapter 1: Introduction) 69 / 97

Chapter 1: Introduction
Lecture Outline
1-1 What is Artificial Intelligence (AI)? 1-4 Applications of AI
1-1-1 AI Definition 1-5 Agents
1-1-2 AI History 1-5-1 Single-Agent Systems vs. Multi-Agent Systems

1-2 Goals of AI 1-5-2 Examples of Agents

1-3 Branches of AI 1-5-3 Environment

1-3-1 Machine Learning (ML)

1-3-2 Natural Language Processing (NLP)

1-3-3 Digital Image Processing & Computer Vision

1-3-4 Robotics

1-3-5 Expert Systems

1-3-6 Evolutionary Algorithms (EA)
1-3-7 Fuzzy Logic
1-3-8 Other Areas

Artificial Intelligence (Chapter 1: Introduction)

1-3-5 Expert Systems
An Expert System is a computer program that simulates the decision-making abilities of a human expert in a specific domain. It uses a
knowledge base of domain-specific information and a set of rules (represented mainly as if–then rules) to solve complex problems that
would typically require human expertise.

In fact, it aims to mimic human reasoning and inference abilities.

Artificial Intelligence (Chapter 1: Introduction) 70 / 97

1-3-5 Expert Systems (continued…)
▌Importance of Expert Systems
● Remember AI’s ultimate goal!!

● Primary Goal: the goal of an expert system is to emulate the decision-making capabilities of a human expert in a particular
domain, providing users with reliable advice, diagnoses, or solutions based on a vast knowledge base.

● Secondary Goals:
 knowledge Dissemination: to distribute expert knowledge and best
practices across organizations or to individuals who do not have direct
access to human experts.
 Decision Support: to assist non-experts in making informed decisions by
providing them with expert-level guidance and recommendations.
 Problem-Solving: to solve complex problems that require specialized
knowledge, often in areas like medicine, engineering, finance, and
customer support.

● Consistency and Objectivity: delivers consistent decisions based on pre-

defined rules and logic, avoiding the variability and biases inherent in
human judgment.

● Access to Expertise: provides access to expert knowledge in areas where

human expertise is scarce, expensive, or unavailable.
Artificial Intelligence (Chapter 1: Introduction) 71 / 97
1-3-5 Expert Systems (continued…)
▌What are the Applications of Expert Systems?

● Medical Diagnosis: systems like MYCIN were developed to diagnose infectious diseases and recommend treatments.
● Financial Services: expert systems are used in stock trading, risk management, loan approval, and fraud detection.
● Customer Support: chatbots and automated support systems that provide users with solutions based on an underlying expert
system.
● Manufacturing: systems that monitor and control complex manufacturing processes, ensuring quality and efficiency.
● Legal Advisory: systems that help in legal research, case analysis, and offering preliminary legal advice.
● Agriculture: expert systems that assist farmers in making decisions
about crop management, pest control, and irrigation.
● Energy Management: systems used in power plants to manage
resources, optimize energy production, and prevent failures.
● etc.

Artificial Intelligence (Chapter 1: Introduction) 72 / 97

1-3-5 Expert Systems (continued…)
Samples of the rules of the expert system

Artificial Intelligence (Chapter 1: Introduction) 73 / 97

Chapter 1: Introduction
Lecture Outline
1-1 What is Artificial Intelligence (AI)? 1-4 Applications of AI
1-1-1 AI Definition 1-5 Agents
1-1-2 AI History 1-5-1 Single-Agent Systems vs. Multi-Agent Systems

1-2 Goals of AI 1-5-2 Examples of Agents

1-3 Branches of AI 1-5-3 Environment

1-3-1 Machine Learning (ML)

1-3-2 Natural Language Processing (NLP)

1-3-3 Digital Image Processing & Computer Vision

1-3-4 Robotics

1-3-5 Expert Systems

1-3-6 Evolutionary Algorithms (EA)
1-3-7 Fuzzy Logic
1-3-8 Other Areas

Artificial Intelligence (Chapter 1: Introduction)

1-3-6 Evolutionary Algorithms (EA)
Evolutionary Algorithms (EAs) are a family of optimization algorithms inspired by the process of natural evolution. They use
mechanisms such as selection, mutation, recombination (crossover), and reproduction to evolve a population of candidate solutions
over successive generations, with the aim of finding the best possible solution to a given problem.

Core Idea: EAs mimic the process of natural selection, where the fittest individuals are more likely to survive and reproduce, passing on
their traits to the next generation. Over time, this leads to the evolution of solutions that are well-suited to the problem at hand.

▌What is the Goal of Evolutionary Algorithms?

● Primary Goal: the primary goal of evolutionary algorithms is to find an optimal or near-optimal solution to complex problems by
iteratively improving a population of solutions based on a fitness function.

Artificial Intelligence (Chapter 1: Introduction) 74 / 97

1-3-6 Evolutionary Algorithms (EA) (continued…)
▌What are the Importances of Evolutionary Algorithms?

● Handling Complexity: evolutionary algorithms are particularly useful for solving complex, non-linear, multi-modal optimization
problems that are difficult for traditional methods to handle.

● Global Search Capability: they have a strong global search capability, which helps in avoiding local optima and finding the global
optimum in challenging problem spaces.

● Flexibility and Adaptability: EAs are highly flexible and can be applied to a wide range of problems, including those with discrete,
continuous, or mixed-type variables. They can also be easily adapted to handle dynamic or multi-objective optimization
problems.

● Robustness: these algorithms are robust in the presence of noisy, imprecise, or incomplete data, making them suitable for real-
world applications where such challenges are common.

Artificial Intelligence (Chapter 1: Introduction) 75 / 97

1-3-6 Evolutionary Algorithms (EA) (continued…)
▌Genetic Algorithms (GA)

Genetic Algorithms (GAs) are a type of evolutionary algorithm that mimics the process of natural selection and genetics to solve
optimization and search problems. They operate by evolving a population of candidate solutions over successive generations to find the
best or most optimal solution.

Core Idea: GAs simulate the process of natural evolution, where the fittest individuals are selected for reproduction. Over many
generations, the population evolves, with better solutions emerging as a result of the selection, crossover, and mutation processes.

Artificial Intelligence (Chapter 1: Introduction) 76 / 97

1-3-6 Evolutionary Algorithms (EA) (continued…)
▌Genetic Algorithms (GA)
The flowchart of GA:

Artificial Intelligence (Chapter 1: Introduction) 77 / 97

1-3-6 Evolutionary Algorithms (EA) (continued…)
▌Genetic Algorithms (GA)
Some concepts in GA:

Selection

Mutation Fitness Calculation

Artificial Intelligence (Chapter 1: Introduction) 78 / 97

1-3-6 Evolutionary Algorithms (EA) (continued…)
▌Genetic Algorithms (GA)
Some applications of GA:
● Circuit Design: Optimizing electronic circuits, such as finding the best layout or component values to achieve desired electrical
characteristics.
● Hyperparameter Tuning in ML: optimizing hyperparameters of machine learning models like neural networks, decision trees, and
support vector machines to enhance model performance.
● Path Planning: finding the most efficient path for a robot to navigate through an environment, avoiding obstacles and minimizing
travel time or energy consumption.
● Trading Strategies: developing and optimizing trading algorithms that make buy/sell decisions in financial markets based on
historical data, technical indicators, and market conditions.
● Drug Discovery: Identifying and optimizing potential drug molecules by simulating their interaction with biological targets,
reducing the time and cost of drug development.
● etc.

Artificial Intelligence (Chapter 1: Introduction) 79 / 97

Chapter 1: Introduction
Lecture Outline
1-1 What is Artificial Intelligence (AI)? 1-4 Applications of AI
1-1-1 AI Definition 1-5 Agents
1-1-2 AI History 1-5-1 Single-Agent Systems vs. Multi-Agent Systems

1-2 Goals of AI 1-5-2 Examples of Agents

1-3 Branches of AI 1-5-3 Environment

1-3-1 Machine Learning (ML)

1-3-2 Natural Language Processing (NLP)

1-3-3 Digital Image Processing & Computer Vision

1-3-4 Robotics

1-3-5 Expert Systems

1-3-6 Evolutionary Algorithms (EA)
1-3-7 Fuzzy Logic
1-3-8 Other Areas

Artificial Intelligence (Chapter 1: Introduction)

1-3-7 Fuzzy Logic
Fuzzy Logic is a form of multi-valued logic that deals with reasoning that is approximate rather than fixed and exact. Unlike traditional
binary logic, which operates on true/false or 0/1 values, fuzzy logic allows for degrees of truth or partial truths. It is based on the
concept of fuzzy sets, where an element's membership in a set is expressed in degrees ranging from 0 to 1.

Core Idea: Fuzzy Logic captures the way humans make decisions in situations of uncertainty, vagueness, and imprecision. It models
real-world complexities by allowing partial membership in sets, which is more aligned with human thinking.

Artificial Intelligence (Chapter 1: Introduction) 80 / 97

1-3-7 Fuzzy Logic (continued…)
The primary goal of Fuzzy Logic is to provide a mathematical framework for dealing with imprecise, vague, or uncertain information. It
aims to model and reason about uncertain or fuzzy concepts in a way that mimics human reasoning.

It also:
 Handle Uncertainty: Enable systems to make decisions and draw conclusions even when input data is noisy, incomplete, or
ambiguous.
 Improve Decision-Making: Enhance decision-making processes in complex environments by incorporating human-like reasoning.
 Control Complex Systems: Design control systems that can manage complex and nonlinear processes without precise
mathematical models.

▌The idea of fuzzy logic was first advanced by Lotfi Aliasker Zadeh, of the University of California at Berkeley
in the 1960s (his father was Rahim Aleskerzade, an Iranian Muslim a journalist from Ardabil).

Artificial Intelligence (Chapter 1: Introduction) 81 / 97

1-3-7 Fuzzy Logic (continued…)
What are the Importances of Fuzzy Logic?

● Handling Ambiguity: fuzzy logic is particularly valuable in situations where the data is uncertain, ambiguous, or incomplete. It allows
for more nuanced and flexible decision-making than traditional binary logic.
● Human-Like Reasoning: fuzzy logic systems can model human reasoning more closely than classical logic systems, making them ideal
for applications where human decision-making processes need to be replicated or improved.
● Robustness in Control Systems: fuzzy logic control systems can be more robust and flexible than traditional control systems,
especially in complex, non-linear, and dynamic environments.
● Ease of Implementation: fuzzy logic is relatively easy to implement and can be applied to a wide range of problems without needing
complex mathematical models or precise data.

Artificial Intelligence (Chapter 1: Introduction) 82 / 97

1-3-7 Fuzzy Logic (continued…)
What are the Applications of Fuzzy Logic?

● Consumer Electronics: Washing machines, cameras, and air conditioners that adjust their operation based on fuzzy logic to improve
user experience. For instance, a fuzzy logic-based washing machine can adjust the washing cycle based on the load's weight,
dirtiness, and fabric type.

● Automotive Systems: Fuzzy logic is used in anti-lock braking systems (ABS), automatic transmission systems, and climate control to
provide smoother and more responsive control.

● Industrial Automation: Fuzzy logic controllers are widely used in industrial processes to manage systems that are too complex for
conventional control methods, such as chemical plant operations or robotics.

● Medical Diagnosis: In medical decision support systems, fuzzy logic helps in diagnosing diseases based on symptoms that are not
strictly defined or where there are overlapping symptoms for multiple conditions.

● Financial Systems: Fuzzy logic is used in credit scoring, risk assessment, and stock market analysis to deal with the uncertainties and
imprecise nature of financial data.

● Expert Systems: Fuzzy logic enhances expert systems by allowing them to handle uncertain or approximate information, which is
often the case in real-world expert decision-making.

● etc.

Artificial Intelligence (Chapter 1: Introduction) 83 / 97

1-3-7 Fuzzy Logic (continued…)
▌Fuzzy Sets
A fuzzy set is an extension of a classical (or crisp) set, which allows for the gradual assessment of the membership of elements in a set.
Unlike classical sets, where an element either belongs or does not belong to a set (i.e., membership is binary: 0 or 1), a fuzzy set allows
for degrees of membership.

In a fuzzy set, an element can belong to the set partially, with a degree of membership ranging between 0 (completely not a member)
and 1 (fully a member). This enables fuzzy sets to represent uncertainty, vagueness, and imprecision more naturally than traditional
binary logic.

Human age groups illustrated as fuzzy sets and membership functions. “Young”, “Middle-Aged”, and “Old” are examples of fuzzy sets. In this illustration, the set “Young” includes
ages from 0 to 37. “µ” represents the value of the membership function. Membership in the set “Young” decreases gradually from ages 23 to 37. An age of 30 has a membership
value µ of 0.5 in the set “Young”. The solid lines represent the fuzzy membership functions for each set. The dashed lines illustrate how “Young” could be made into a crisp, non-
fuzzy set. In this illustration, the crisp, non-fuzzy set “Young” includes ages from 0 to 30.

Artificial Intelligence (Chapter 1: Introduction) 84 / 97

Chapter 1: Introduction
Lecture Outline
1-1 What is Artificial Intelligence (AI)? 1-4 Applications of AI
1-1-1 AI Definition 1-5 Agents
1-1-2 AI History 1-5-1 Single-Agent Systems vs. Multi-Agent Systems

1-2 Goals of AI 1-5-2 Examples of Agents

1-3 Branches of AI 1-5-3 Environment

1-3-1 Machine Learning (ML)

1-3-2 Natural Language Processing (NLP)

1-3-3 Digital Image Processing & Computer Vision

1-3-4 Robotics

1-3-5 Expert Systems

1-3-6 Evolutionary Algorithms (EA)
1-3-7 Fuzzy Logic
1-3-8 Other Areas

Artificial Intelligence (Chapter 1: Introduction)

1-3-8 Other Areas
There are several other important areas within AI that extend beyond the fields that are mentioned before. Here are additional key
areas in AI:

▌Deep Learning
is a subset of machine learning that involves neural networks with many layers (hence "deep") to model and solve complex problems. These
networks are designed to automatically learn representations of data through hierarchical layers, where each layer extracts increasingly
abstract features.

▌Knowledge Representation and Reasoning (KRR)

is a branch of AI that focuses on how to represent information about the world in a form that a computer system can understand and use to
solve complex problems. It also involves reasoning, or using the knowledge to draw conclusions, make decisions, or infer new information.

▌Cognitive Computing
refers to systems that simulate human thought processes to help solve complex problems. These systems use AI techniques like machine
learning, natural language processing, and reasoning to mimic the functioning of the human brain’s cognitive functionality. Cognitive
computing aims to create systems that can learn, reason, interact naturally, and make decisions based on large amounts of unstructured data.

Artificial Intelligence (Chapter 1: Introduction) 85 / 97

1-3-8 Other Areas (continued…)
There are several other important areas within AI that extend beyond the fields that are mentioned before. Here are additional key
areas in AI:

▌Multi-Agent Systems (MAS)

refer to systems composed of multiple interacting intelligent agents, which can be either software-based or physical entities. These agents
work together or compete to solve problems that are beyond the capability of an individual agent. MAS are designed for distributed
environments where agents can communicate, collaborate, and make decisions independently or collectively.

▌AI in Healthcare
refers to the use of artificial intelligence technologies to improve medical care, from diagnosis and treatment to patient management and
drug discovery. AI systems can analyze large datasets, detect patterns, and provide insights that assist healthcare professionals in making
more accurate decisions.

▌Bioinformatics
is an interdisciplinary field that combines biology, computer science, and information technology to analyze and interpret biological data. It
involves the development and application of computational tools and algorithms to understand complex biological systems, such as DNA
sequencing, protein structure prediction, and gene expression analysis.

Artificial Intelligence (Chapter 1: Introduction) 86 / 97

1-3-8 Other Areas (continued…)
There are several other important areas within AI that extend beyond the fields that are mentioned before. Here are additional key
areas in AI:

▌Quantum Computing and AI

refers to the integration of quantum computing principles with AI to enhance the capabilities of AI algorithms. Quantum computers use
quantum bits (qubits) and quantum principles like superposition and entanglement, allowing them to process complex information faster
than classical computers.

▌AI in Cybersecurity
involves using artificial intelligence technologies to detect, prevent, and respond to cyber threats. AI systems can analyze large volumes of
data, identify patterns, and detect anomalies in real-time, allowing for faster and more accurate threat detection than traditional methods.

▌Swarm Intelligence
is a concept in artificial intelligence inspired by the collective behavior of decentralized, self-organized systems, often observed in nature
among groups of animals like ants, bees, or birds. It involves designing algorithms that mimic these natural behaviors to solve complex
optimization and coordination problems.

▌etc.

Artificial Intelligence (Chapter 1: Introduction) 87 / 97

Chapter 1: Introduction
Lecture Outline
1-1 What is Artificial Intelligence (AI)? 1-4 Applications of AI
1-1-1 AI Definition 1-5 Agents
1-1-2 AI History 1-5-1 Single-Agent Systems vs. Multi-Agent Systems

1-2 Goals of AI 1-5-2 Examples of Agents

1-3 Branches of AI 1-5-3 Environment

1-3-1 Machine Learning (ML)

1-3-2 Natural Language Processing (NLP)

1-3-3 Digital Image Processing & Computer Vision

1-3-4 Robotics

1-3-5 Expert Systems

1-3-6 Evolutionary Algorithms (EA)
1-3-7 Fuzzy Logic
1-3-8 Other Areas

Artificial Intelligence (Chapter 1: Introduction)

1-4 Applications of AI
In today's rapidly advancing technological landscape, AI has become an integral force, transforming and governing various aspects of
our lives. From revolutionizing industries to enhancing everyday conveniences, AI is continually pushing boundaries and reshaping the
way we interact with the world. Its impact is evident in numerous sectors, where it not only streamlines processes and solves complex
problems but also opens new avenues for innovation. As AI technology evolves, its applications expand, influencing everything from
how we communicate to how we make decisions and solve global challenges.

Lets to investigate some applications of AI….

Artificial Intelligence (Chapter 1: Introduction) 88 / 97

1-4 Applications of AI (continued…)

■ Healthcare: AI in healthcare includes diagnostic tools that analyze medical images, predictive models for disease outbreaks,
personalized treatment plans, and virtual health assistants. These applications help in improving diagnostic accuracy, tailoring
treatments to individual patients, and enhancing overall healthcare delivery.

■ Finance: in the financial sector, AI is used for fraud detection, algorithmic trading, risk assessment, and personalized financial advice.
AI systems analyze transaction patterns to detect anomalies, optimize trading strategies, and provide personalized investment
recommendations.

■ Transportation: AI is a key driver of advancements in autonomous vehicles, traffic management systems, and logistics optimization.
Self-driving cars use AI to navigate and make decisions, while AI algorithms optimize routing and delivery processes for logistics
companies.

■ Entertainment: in entertainment, AI powers recommendation engines for streaming services, generates content through tools like
deepfake technology, and enhances user experiences in video games with intelligent NPCs (non-player characters) and adaptive
difficulty levels.

■ Education: AI in education includes personalized learning platforms, intelligent tutoring systems, and administrative automation.
These applications help tailor educational experiences to individual student needs, automate grading, and provide real-time
feedback.

Artificial Intelligence (Chapter 1: Introduction) 89 / 97

1-4 Applications of AI (continued…)

■ Manufacturing: AI is used in predictive maintenance, quality control, and production optimization. Machine learning algorithms
analyze sensor data to predict equipment failures, while AI-powered vision systems detect defects in manufacturing processes.

■ Agriculture: In agriculture, AI aids in precision farming, crop monitoring, and livestock management. AI-driven systems analyze data
from drones and sensors to optimize planting strategies, monitor crop health, and manage resources more efficiently.

■ Cybersecurity: AI enhances cybersecurity through real-time threat detection, automated incident response, and vulnerability
management. AI systems analyze network traffic to identify and respond to potential threats and anomalies.

■ Retail: AI enhances the retail experience through personalized recommendations, inventory management, and customer service.
Chatbots and virtual assistants provide real-time support, while AI-driven analytics help retailers understand customer preferences
and optimize supply chains.

■ Smart Cities: AI contributes to the development of smart cities by optimizing energy usage, managing traffic flow, and improving
public services. AI-powered systems monitor and analyze data from various city infrastructure components to enhance urban living
conditions.

■ Human Resources: In HR, AI streamlines recruitment processes through resume screening, candidate matching, and interview
scheduling. AI also aids in employee performance evaluation and engagement through data-driven insights.

■ etc.
Artificial Intelligence (Chapter 1: Introduction) 90 / 97
Chapter 1: Introduction
Lecture Outline
1-1 What is Artificial Intelligence (AI)? 1-4 Applications of AI
1-1-1 AI Definition 1-5 Agents
1-1-2 AI History 1-5-1 Single-Agent Systems vs. Multi-Agent Systems

1-2 Goals of AI 1-5-2 Examples of Agents

1-3 Branches of AI 1-5-3 Environment

1-3-1 Machine Learning (ML)

1-3-2 Natural Language Processing (NLP)

1-3-3 Digital Image Processing & Computer Vision

1-3-4 Robotics

1-3-5 Expert Systems

1-3-6 Evolutionary Algorithms (EA)
1-3-7 Fuzzy Logic
1-3-8 Other Areas

Artificial Intelligence (Chapter 1: Introduction)

1-5 Agents
In the context of artificial intelligence, an agent is an autonomous entity that perceives its environment (through sensors) and acts
upon that environment (through actuators) to achieve specific goals. Agents can be software programs, robots, or other systems
capable of taking actions based on their perceptions and internal decision-making processes.

Agents are designed to operate in dynamic environments where they can make decisions, solve problems, and adapt to changes in
their surroundings.

Artificial Intelligence (Chapter 1: Introduction) 91 / 97

1-5 Agents (continued…)
Agents, are entities that exhibit certain properties, characteristics, and behaviors. These properties define the nature and
functionality of agents within the system. Here are “some” key properties of an agent in MAS:

1. Autonomy: Agents in MAS are typically designed to be autonomous, meaning they have the ability to operate independently,
make decisions based on their own perceptions, and take actions to achieve their goals without constant external control.

2. Perception: Agents have the capability to perceive their environment using sensors or other means. Perception allows agents
to gather information about the state of the environment, the actions of other agents, and relevant changes.

3. Decision-Making: Agents possess decision-making capabilities. They can evaluate their perceptions, internal knowledge, and
goals to make decisions on how to act in a given situation. The decision-making process is a crucial aspect of an agent's
intelligence.

4. Goal-Directed Behavior: Agents are typically designed with specific goals or objectives. Goal-directed behavior motivates
agents to take actions that move them closer to achieving their objectives.

5. Communication (social ability): Agents may communicate with other agents within the system. Communication enables
agents to share information, coordinate actions, negotiate, and collaborate. Effective communication is essential for achieving
collective goals.

Artificial Intelligence (Chapter 1: Introduction) 92 / 97

Chapter 1: Introduction
Lecture Outline
1-1 What is Artificial Intelligence (AI)? 1-4 Applications of AI
1-1-1 AI Definition 1-5 Agents
1-1-2 AI History 1-5-1 Single-Agent Systems vs. Multi-Agent Systems

1-2 Goals of AI 1-5-2 Examples of Agents

1-3 Branches of AI 1-5-3 Environment

1-3-1 Machine Learning (ML)

1-3-2 Natural Language Processing (NLP)

1-3-3 Digital Image Processing & Computer Vision

1-3-4 Robotics

1-3-5 Expert Systems

1-3-6 Evolutionary Algorithms (EA)
1-3-7 Fuzzy Logic
1-3-8 Other Areas

Artificial Intelligence (Chapter 1: Introduction)

1-5-1 Single-Agent Systems vs. Multi-Agent Systems
Single-Agent Systems: a Single-Agent System involves one autonomous agent that interacts with its environment to achieve specific
goals. This agent operates independently without direct interaction or collaboration with other agents.

Multi-Agent Systems: a Multi-Agent System (MAS) involves multiple interacting agents that work together or compete to achieve their
individual or collective goals. These systems are designed to solve problems that are too complex for a single agent.

Artificial Intelligence (Chapter 1: Introduction) 93 / 97

Chapter 1: Introduction
Lecture Outline
1-1 What is Artificial Intelligence (AI)? 1-4 Applications of AI
1-1-1 AI Definition 1-5 Agents
1-1-2 AI History 1-5-1 Single-Agent Systems vs. Multi-Agent Systems

1-2 Goals of AI 1-5-2 Examples of Agents

1-3 Branches of AI 1-5-3 Environment

1-3-1 Machine Learning (ML)

1-3-2 Natural Language Processing (NLP)

1-3-3 Digital Image Processing & Computer Vision

1-3-4 Robotics

1-3-5 Expert Systems

1-3-6 Evolutionary Algorithms (EA)
1-3-7 Fuzzy Logic
1-3-8 Other Areas

Artificial Intelligence (Chapter 1: Introduction)

1-5-2 Examples of Agents
Examples of Agents
● Personal Digital Assistants (PDAs): like Siri, Google Assistant, or Amazon Alexa. These are software agents that interact with users to
perform tasks such as setting reminders, answering questions, and managing schedules based on voice commands.

● Autonomous Vacuum Cleaners: like Roomba. These robotic vacuum cleaners navigate and clean floors independently using sensors
to detect obstacles and dirt. They make decisions on their cleaning paths based on pre-programmed algorithms.

● Chatbots: like customer service chatbots on websites. These are software agents designed to simulate conversation with users,
answer queries, and provide support based on predefined responses and natural language processing.

● Recommendation Systems: like Netflix or Amazon recommendation engines. These systems analyze user behavior and preferences
to suggest movies, products, or content, operating as a single-agent system to personalize recommendations.

● Autonomous Drones: like aerial photography drones. These drones are capable of flying and capturing images or video based on
programmed routes or user commands, operating autonomously in their task.

Artificial Intelligence (Chapter 1: Introduction) 94 / 97

1-5-2 Examples of Agents (continued…)
Examples of Agents
● Autonomous Vehicle Fleets: like a fleet of self-driving taxis. Multiple autonomous vehicles coordinate with each other to provide
ride-sharing services, optimize routes, and manage traffic efficiently within a city.

● Traffic Management Systems: like smart traffic lights and sensors network. Multiple agents (traffic lights, sensors, and control
systems) interact to manage traffic flow, reduce congestion, and improve road safety.

● Robotic Swarms: like swarms of drones used for environmental monitoring. A group of drones work together to cover large areas,
gather data, and coordinate their actions for tasks like search and rescue or surveillance.

● Market Simulation Systems: like stock market trading simulations. Multiple trading agents (both human and algorithmic) interact in a
simulated financial market, making trades and responding to market conditions based on strategies and signals.

● Collaborative Robotics: like multi-robot systems in manufacturing.

Robots in a manufacturing environment work together on
assembly lines or production tasks, coordinating their actions to
optimize efficiency and productivity.

Artificial Intelligence (Chapter 1: Introduction) 95 / 97

Chapter 1: Introduction
Lecture Outline
1-1 What is Artificial Intelligence (AI)? 1-4 Applications of AI
1-1-1 AI Definition 1-5 Agents
1-1-2 AI History 1-5-1 Single-Agent Systems vs. Multi-Agent Systems

1-2 Goals of AI 1-5-2 Examples of Agents

1-3 Branches of AI 1-5-3 Environment

1-3-1 Machine Learning (ML)

1-3-2 Natural Language Processing (NLP)

1-3-3 Digital Image Processing & Computer Vision

1-3-4 Robotics

1-3-5 Expert Systems

1-3-6 Evolutionary Algorithms (EA)
1-3-7 Fuzzy Logic
1-3-8 Other Areas

Artificial Intelligence (Chapter 1: Introduction)

1-5-3 Environment
In multi-agent systems, an environment is a shared space where agents can interact with each other and with their surroundings. The
environment encompasses the surroundings, conditions, and factors that affect the behavior and interactions of the agents.

Understanding the environment is crucial in designing and analyzing multi-agent systems, as it directly influences the agents'
perceptions, actions, and overall system dynamics.

Artificial Intelligence (Chapter 1: Introduction) 96 / 97

1-5-3 Environment (continued…)
Let to consider some related concepts to environment in MAS:

▌Definition: The environment in a multi-agent system encompasses everything external to individual agents. It includes the
space, conditions, and factors that influence the agents' actions and perceptions.

▌Perception: Agents perceive the environment through sensors or observation mechanisms. The information gathered from the
environment forms the basis for the agents' decision-making processes.

▌State: The state of the environment represents the current configuration or condition of relevant variables. Changes in the
environment's state impact the perceptions and interactions of the agents.

▌Action Space: The environment defines the set of possible actions that agents can take. Agents interact with the environment
by executing actions, and these actions may lead to changes in the state of the environment.

Understanding the environment is essential for designing effective multi-agent systems. It influences the development of agent
architectures, communication protocols, and coordination mechanisms.

Artificial Intelligence (Chapter 1: Introduction) 97 / 97

█ Exercises
Try to find appropriate responses for the following questions.

1-1 ▌ Define in your own words: (i) intelligence, (ii) artificial intelligence, (iii) agent, (iv) rationality, (v) logical reasoning.

1-2 ▌ Using your own words try to describe the goal of AI.

1-3 ▌ If you were asked why AI is divided into branches, how would you respond?

1-4 ▌ Using your own words try to describe “concept learning”.

1-5 ▌ (i) What is the relation between “artificial neural networks” and “machine learning”? (ii) Do all machine learning methods require artificial
neural networks

1-6 ▌ What is the role of “data” in ML (and generally in concept learning)?

1-7 ▌ (i) What are supervised and unsupervised learning? (ii) How do they differ from each other? (iii) Are there other types of machine learning
algorithms? (iv) Try to list well-known algorithms in each category.

1-8 ▌ In your own words: (i) Try to explain the main objectives of each branch of AI, and (ii) List several applications for each.

1-9 ▌ Which sciences play a role in AI?

1-10 ▌ (i) What is an agent in AI? (ii) What are "single-agent" and "multi-agent systems"? (iii) What are the characteristics of agents?

Artificial Intelligence (Chapter 1: Introduction)

█ Exercises
1-11 ▌ There is a kind of learning which is called “Reinforcement Learning (LR)”. Provide a comprehensive and conceiving description about it.

1-12 ▌ What are the specifications of an appropriate dataset for learning a model?

1-13 ▌ To what extent are the following computer systems instances of artificial intelligence:
- Supermarket bar code scanners.
- Voice-activated telephone menus.
- Spelling and grammar correction features in Microsoft Word.
- Internet routing algorithms that respond dynamically to the state of the network.

1-14 ▌ Is AI a science, or is it engineering? Or neither or both? Explain.

1-15 ▌ “Surely computers cannot be intelligent—they can do only what their programmers tell them.” Is the latter statement true, and does it imply the
former?

1-16 ▌ “Surely animals cannot be intelligent—they can do only what their genes tell them.” Is the latter statement true, and does it imply the former?

1-17 ▌ “Surely animals, humans, and computers cannot be intelligent—they can do only what their constituent atoms are told to do by the laws of
physics.” Is the latter statement true, and does it imply the former?

Artificial Intelligence (Chapter 1: Introduction)