Kuliah Pengantar Kecerdasan Buatan oleh Dr. Sunu Wibirama. Kuliah ini dibagi menjadi empat bagian, yakni:
Part 1: Revolusi Industri 4.0 dan Kecerdasan Buatan
Part 2: Sejarah Turing Machine dan Teknologi Kecerdasan Buatan
Part 3: Pengantar Machine Learning
Part 4: Pengantar Deep Neural Network
Instruktur:
Dr. Sunu Wibirama (UGM, Indonesia)
http://sunu.staff.ugm.ac.id
Introduction to Artificial Intelligence - Pengenalan Kecerdasan Buatan
1. www.ugm.ac.idLocally Rooted, Globally Respected
Artificial Intelligence – Department of Electrical Engineering and Information Technology Dr. Sunu Wibirama
Introduction to Artificial Intelligence
Dr. Sunu Wibirama
sunu@ugm.ac.id
Department of Electrical Engineering
and Information Technology
Faculty of Engineering
Universitas Gadjah Mada
Short profile
▪ Dr. Sunu Wibirama
Teknik Elektro, Universitas Gadjah Mada
Yogyakarta, Indonesia (S.T.) – 2007
Dept. Electronics, King Mongkut’s Institute
of Technology Ladkrabang, Bangkok,
Thailand (M.Eng.) – 2010
Graduate School of Science and Technology,
Tokai University, Tokyo, Japan
(Dr.Eng.) – 2014
sunu@ugm.ac.id
http://sunu.staff.ugm.ac.id
http://bit.ly/gazetracking
Affiliation:
Intelligent Systems RG
Dept. of Electrical Engineering &
Information Technology
Faculty of Engineering
Universitas Gadjah Mada
Area of research
• eye-gaze tracking
• computer vision
• human-computer interaction
• user experience
• human factors and safety in 3D
technology and virtual reality
2. http://sunu.staff.ugm.ac.id
Goal of this introductory session
• Understanding basic concept of artificial intelligence (AI)
• Differentiating AI with machine learning and deep learning
• Identifying various applications of AI that support daily activities
• Pointing out some milestones in history of AI
• Comparing basic differences of machine learning and deep learning
3. Overview of the lecture
• Part 1: Industrial Revolution 4.0 and Artificial Intelligence
• Part 2: History of Turing Machine and Current Status of Artificial Intelligence
• Part 3: Introduction to Machine Learning
• Part 4: The Rise of Deep Neural Network
The Fourth Industrial
Revolution
6. Artificial intelligence (AI)
• Intelligence: the ability to acquire and apply knowledge
• Artificial intelligence is created to simulates human intelligence processes
by machines, especially computer systems.
• These processes include learning, decision-making, and self-correction.
Particular applications of AI include expert systems (e.g., Google Maps),
speech recognition (e.g.,: Apple Siri) and machine vision
(e.g., Facebook’s face recognition).
Can you find one example of AI
technology that you encounter in
your life?
Short quiz
7. End of Part 1
www.ugm.ac.idLocally Rooted, Globally Respected
Artificial Intelligence – Department of Electrical Engineering and Information Technology Dr. Sunu Wibirama
Introduction to Artificial Intelligence
(part 2)
Dr. Sunu Wibirama
sunu@ugm.ac.id
Department of Electrical Engineering
and Information Technology
Faculty of Engineering
Universitas Gadjah Mada
9. AI - some years ago
Alan Turing has developed
a Turing Test in 1950
while working in
University of Manchester, UK.
The "standard interpretation" of the
Turing test, in which player C (the
interrogator) is given the task of
trying to determine which player – A
or B – is a computer and which is a
human. The interrogator is limited to
using the responses to written
questions to make the
determination
AI - some years ago
IBM's Watson beat two Jeopardy champions
in a special edition of game show in 2011
10. AI - Google I/O2018
What do you think?
Has AI passed the Turing Test?
Short quiz
11. End of Part 2
www.ugm.ac.idLocally Rooted, Globally Respected
Artificial Intelligence – Department of Electrical Engineering and Information Technology Dr. Sunu Wibirama
Introduction to Artificial Intelligence
(part 3a)
Dr. Sunu Wibirama
sunu@ugm.ac.id
Department of Electrical Engineering
and Information Technology
Faculty of Engineering
Universitas Gadjah Mada
16. Now, what is this?
Human can make an inference almost
effortlessly, but you cannot expect the
same thing on computer.
We train a computer with training data
and we expect the computer to make
inference over new (test) data.
How machine learning works (simplified)
17. Machine Learning for
Vehicle Plate Number Recognition
DIP DIP
DIP
Researcher:
D.Sihombing, H.A. Nugroho, S. Wibirama (2015)
A. Prasetyo, S. Wibirama, N.A. Setiawan (2016)
Machine learning
Machine learning for traffic monitoring
Researcher:
D.A. Kurniawan, S. Wibirama, N.A. Setiawan (2016)
18. Machine learning in medical eye tracking
San Diego, US
Industrial partner:
Artificial Intelligence in Autonomous Navigation
DARPA Grand Challenge
Stanley autonomous car
Google intelligent car
19. Let's see a video about a mind
blowing self-driving car....
20. If AI can drive a car, do you think that
in future we do not need human
drivers for public transports anymore?
Short quiz
End of Part 3
23. Neural networks was used to recognize
hand writing in a bank check (LeCun,1999)
The Hype is Real for Sure...
Even the Twilight's girl has published a Deep Learning paper
24. The Rise of Deep Learning in Silicon Valley
• 2012 was the first year that neural nets grew to prominence
• Alex Krizhevsky, Ilya Sutskever, and Geoffrey Hinton used them to win that
year’s ImageNet competition (basically, it is the annual olympics of computer
vision)
• They dropped the classification error record from 26.2% to 15.3%,
a remarkable improvement at the time.
• Ever since then, a host of companies have been using deep learning at the
core of their services:
• Facebook uses neural nets for their automatic tagging algorithms
• Google for their photo search
• Amazon for their product recommendations
• Pinterest for their home feed personalization
• Instagram for their search infrastructure.
Mind-blowing citations in 6 years
21645 papers have cited them since 2012
25. To whom you can refer
Dr. Yann LeCun
Head, Facebook AI Lab
Prof. Geoffrey Hinton
Canadian Institute
for Advanced Research /
Google Brain
Prof. Yoshua Bengio
Montreal Institute
for Learning Algorithm
Dr. Andrew Ng
Coursera, Baidu,
Stanford University
pic of andrew ng
https://www.cifar.ca/research/learning-in-machine-and-brains/
Canadian Institute for Advanced Research (CIFAR)
33. Pros and cons
Source: Introducing Deep Learning with
Matlab (Mathworks, 2018)
Summary
• Artificial intelligence has been used in so many applications. Practically,
we are surrounded by AI technologies, in so many forms.
• Machine learning is a sub-field of AI that focuses on giving ability to a
computer to learn from data without explicitly being programmed by a
human.
• Deep-learning is a new form of neural network research. It is now the most
popular algorithm of machine learning technology used in various
companies.