Applied Artificial

Intelligence

Dr. Huseyin Kusetogullari

 Fuzzy Logic
• Fuzzy logic is a mathematical approach to
reasoning that deals with imprecise or uncertain
information, mimicking how humans make
decisions in real-world situations.

• Unlike classical logic, which operates on binary

true/false values (0 or 1), fuzzy logic allows for
degrees of truth, where something can partially
belong to multiple sets or categories.
Department of Computer Science and
Engineering, Blekinge Institute of 2
Technology (BTH).
 Key Concepts in Fuzzy Logic:
• Fuzzy Sets: In classical set theory, an element
either belongs to a set or doesn't (e.g., a value is
either "hot" or "cold").

• In fuzzy logic, an element can belong to multiple

sets to different degrees.

• For example, the temperature could be

considered both "warm" and "hot" at the same
time, with certain degrees of membership in both
categories
Department of Computer Science and
Engineering, Blekinge Institute of 3
Technology (BTH).
 Membership Function
• Membership Function: This defines how much an
element belongs to a fuzzy set.

• The membership degree is a value between 0 and

1, where 0 means no membership and 1 means
full membership.

• Example: If the temperature is 28°C, it could have

a membership of 0.6 in the "warm" set and 0.4 in
the "hot" set.

Department of Computer Science and

Engineering, Blekinge Institute of 4
Technology (BTH).
Cont.
• Linguistic Variables: Fuzzy logic uses linguistic
terms, like "low," "medium," or "high," to
represent variables.

• These terms are modeled as fuzzy sets, and

the degree to which a value belongs to one of
these sets is determined by the membership
function.
Department of Computer Science and
Engineering, Blekinge Institute of 5
Technology (BTH).
 Fuzzy Rules
• Fuzzy Rules: These are "if-then" statements
that govern the behavior of a fuzzy system.
They are used to handle decisions or
inferences based on the degrees of truth from
the fuzzy sets.
• Example:
• IF the temperature is "warm" AND the humidity is
"high," THEN set the fan speed to "medium."

Department of Computer Science and

Engineering, Blekinge Institute of 6
Technology (BTH).
Cont.
• Fuzzification: This is the process of converting crisp,
accurate and detailed input values (e.g., temperature =
28°C) into fuzzy sets with degrees of membership.

• Defuzzification: After the fuzzy inference process, the

resulting fuzzy set is converted back into a single crisp
value that can be used for practical decision-making.

• This is necessary because, in real-world applications,

actions often require definite values (e.g., fan speed =
70%).
Department of Computer Science and
Engineering, Blekinge Institute of 7
Technology (BTH).
 Application of Fuzzy Logic:
Fuzzy logic is widely used in systems that require handling of
vague or imprecise data, such as:

• Control Systems: For instance, in washing machines and air

conditioners, fuzzy logic controls the machine's behavior
based on uncertain inputs (e.g., temperature, load size).

• Decision-making Systems: Used in AI systems where

decisions have to be made in uncertain environments (e.g.,
autonomous vehicles, robotics).

• Pattern Recognition: For image processing, facial

recognition, etc.
Department of Computer Science and
Engineering, Blekinge Institute of 8
Technology (BTH).
 Example: Fuzzy Logic System for Fan
Speed Control
• Let's apply this algorithm to a fuzzy logic
control system for fan speed based on
temperature and humidity.

Department of Computer Science and

Engineering, Blekinge Institute of 9
Technology (BTH).
Department of Computer Science and
Engineering, Blekinge Institute of 10
Technology (BTH).
Department of Computer Science and
Engineering, Blekinge Institute of 11
Technology (BTH).
Fuzzification

Department of Computer Science and

Engineering, Blekinge Institute of 12
Technology (BTH).
Department of Computer Science and
Engineering, Blekinge Institute of 13
Technology (BTH).
Department of Computer Science and
Engineering, Blekinge Institute of 14
Technology (BTH).
Department of Computer Science and
Engineering, Blekinge Institute of 15
Technology (BTH).
 • Questions?

Department of Computer Science and

Engineering, Blekinge Institute of 16
Technology (BTH).
Assignments and Examination
• You have had;
– 3 Assignments for DV2618 (Assignment 1, 2 and 3)
– 2 Assignments for DV2619 (Assignment 1 and 2)

– IMPORTANT
– First Resubmission will be available in January,
2025
– Second Resubmission will be available in May,
2025
17
About Lectures
• There are
– 10 lectures (on Campus)
– 3 Labs

18
In this course
• Introduction to Applied AI,
• Knowledge representation in AI,
• Graphs and Search,
• Genetic Algorithm,
• Data Mining and Knowledge Discovery,
• Machine learning algorithms
• Fundamentals of Neural Networks and Deep Learning,
• The ethics of AI, Future of AI and Benefits, Challenges,
• Key Applications: Computer Vision, & Natural Language
Processing,

19
Purpose of this course
• Learn how to use the basic search methods;
• Understand the basic methods for problem
formulation and knowledge representation;
• Understand the basic idea of automatic
reasoning;
• Know some basic concepts related to
optimization, data science, prediction, pattern
recognition and machine learning.

20
Recommended research papers
• You can find recommended research papers
on the course webpage in Canvas.

21
What is Artificial Intelligence?

22
23
Artificial Intelligence
• Views of AI fall into four categories (System):
Acting Humanly Acting Rationally
Thinking Humanly Thinking Rationally

* Systems that think like humans.

* Systems that act like humans.
* Systems that think rationally.
* Systems that act rationally.

24
Turing’s Test

25
Turing’s Test

▪ Test proposed by Alan Turing in 1950

▪ The computer is asked questions by a human interrogator. It passes
the test if the interrogator cannot tell whether the responses come
from a person
▪ No physical interaction

https://academic.oup.com/mind/article/LIX/236/433/986238

26
 Systems that act like humans
Computer Human

? human interrogator

• You enter a room which has a computer terminal. You have

a fixed period of time to type what you want into the
terminal, and study the replies. At the other end of the line
is either a human being or a computer system.

• If it is a computer system, and at the end of the period you

cannot reliably determine whether it is a system or a
human, then the system is deemed to be intelligent.
Overview, History and Evaluation

28
29
AI in Big Data
• Facebook
• Youtube: Videos
• Instagram: Images
• Twitter: words, sentences etc. Categorical Data: words

Numerical Data Visual data: images and Videos

Demonstration
Challenges in Data
• There are many challenges for the future in big
data management and analytics.
– Pre-processing,
– Analytics architecture,
– Evaluation,
– Imbalance,
– Anomaly Detection,
– Developing learning methods,
– Time-evolving data,
– Compression,
– Visualization, etc.

32
Motivation
• AI Algorithms are needed for tasks which are too complex for humans.
– Impractical
– Time consuming
– Hard to solve

• The size and complexity of these data sets means that humans are unable to extract useful
information from them.

• For a large amount of data, an AI algorithm will work out by exploring that data to predict
the result.

• Data is everywhere in abundant amounts.

• For instance:
– Surveillance cameras continuously capture video,
– Most financial transactions are recorded
– Satellites generate tera-bytes of data every year
– Digitized written texts from our libraries etc.

• The main purposes of applying AI in human life are to make easier, safer and better.

33
Motivation
• The challenging is to do something useful with this data.
– If the banks learn about spending patterns, can they detect
credit card fraud quickly?
– If hospitals share data, then can treatments that do not work as
well as expected be identified quickly?
– Can an intelligent car give you early warning of problems so that
you do not end up stranded in the worst part of town?

• These are some of the questions that AI methods may be

used to answer.

• Can human answer these questions too?

34
Utilization of AI in NLP and Computer Vision
Basics of Supervised and
Unsupervised Learning

36
 Some AI Algorithms
Theory and practical examples in:
➢Graph Search algorithms (BFS, DFS)
➢Genetic Algorithm (nature-inspired optimization
algorithm)
➢Linear Regression Model
➢Fundamentals of Artificial Neural Networks and
Deep Learning
➢Fuzzy Logic

37
 About Written Exam
• It will be on campus
• It will be a computer –based exam (Inspera)

• About Questions
✓ No coding
✓ The questions will be related to the contents taught during
the course
✓ There can be simple calculations and you sould bring your
own calculator (bring your own calculator, mobile phones
are NOT allowed to be used as calculator)
✓ The exam will be 31st October

38
Thank you for listening!!!

39