You are on page 1of 49

A Beginner’s

Guide to
Machine Learning

Ahmad Ataka Awwalur Rizqi, S.T., Ph.D.


Universitas Gadjah Mada, PIJAR Foundation
About me
• 2014: Co-founded ATNIC, Now JALA Tech
• Jan 2015-April 2019: PhD Robotics @ KCL
• Nov 2016-Apr 2020: Tutor @ Richer Education
• Nov 2017: Co-Founded Jago Robotika
• Jan 2019-Apr 2020: Post Doc @ QMUL
• Mei-Jul 2020: Engineer @ JALA Tech
• Aug-Dec 2020: Researcher at DTETI UGM
• January-Aug 2021: Research Fellow at NTU-
Continental Corporate Lab, Singapore
• March 2021-Now: Technical Advisor @ ABA Robotics
• Aug 2021-Now: Lecturer at DTETI UGM

More details at:


ahmadataka.bitbucket.io
Outline “A begineer’s guide
to machine learning”
1. Machine Learning 101
2. Machine Learning dan Artificial Neural Network
3. Using My First Machine Learning Hands-on
4. The future of Machine Learning in the world and
Indonesia

DISCLAIMER: This course is an introduction course to give you a


basic idea on machine learning. To master the technical detail,
further course is required!
CONFIDENTIAL

3
1. Machine Learning 101
A beginner’s guide to machine learning #1
Presented by
Ahmad Ataka Awwalur Rizqi, S.T., Ph.D.
Universitas Gadjah Mada, PIJAR Foundation
Outline
Machine Learning 101

• Apa itu pembelajaran mesin?


• Sejarah singkat
• Kisah sukses pembelajaran mesin

5
What comes to your mind when hearing
the word “AI”?

6
What comes to your mind when hearing
the term “machine learning”?

7
Apa itu pembelajaran mesin?

“ A field of study that gives computers the


ability to learn without being explicitly
programmed.”

Attributed to Arthur Samuel

Machine Learning =/= Artificial Intelligence


AI involves a machine processing information
from its surrounding environment and then
factoring that information into decisions toward
achieving some desired outcome

8
Sejarah Singkat

9
Kisah sukses AI
dan Pembelejaran Mesin

Text Processing Speech Processing Image Processing

1
0
Kisah sukses - Duolinggo
Fitur Fitur Birdbrain:
Spaced Repetition Machine Learning based

Duolingo builds an exceptionally detailed Birdbrain adjusts the difficulty within a


profile based on what you know and what lesson based on how hard a
you don’t know. sentence is for you specifically

1
1
Kisah sukses - Duolinggo
Fitur Active Learning: Fitur
NLP based Bandit Algorithm

For some mistakes that you make, Duolingo Duolingo even uses AI to improve the
tries to figure out the root cause so it can effectiveness of its notifications. The
offer you a timely tip. app sends you a notification each day
to remind you to practice

1
2
Kisah sukses - Duolinggo
Fitur Fitur
Logistic Regression Chatbots

Whenever you submit an answer to a challenge and Duolingo


says you got it wrong, you have the option to hit the Report
button. If you think you got it right, you can appeal.

Chatbot in 94 languages

1
3
Kisah sukses - Duolinggo
Fitur
CEFR checker

Duolingo used unsupervised machine learning to build a tool for determining the difficulty of any
text for language learners. The team used the Common European Framework of
Reference (CEFR), which has a six-level scale: A1 and A2 (beginner), B1 and B2 (intermediate),
and C1 and C2 (advanced).

1
4
Recap Machine Learning 101

• Machine learning memberikan mesin


kemampuan untuk belajar tanpa harus
diprogram secara eksplisit
• Machine learning tidak sama dengan artificial
intelligence, tetapi merupakan subset/bagian
darinya
• Secara konseptual, bidang AI diawali oleh Alan
Turing
• Di dunia modern, machine learning dan AI
digunakan di berbagai bidang seperti text
processing, image processing, atau speech
processing

1
5
References

• Andrew Trask. 2019. Grokking Deep Learning (1st. ed.).


Manning Publications Co., USA.
• Jon Krohn, Grant Beyleveld, and Aglae Bassens. 2019.
Deep Learning Illustrated: A Visual, Interactive Guide to
Artificial Intelligence (1st. ed.). Addison-Wesley
Professional.
• Materi Microcredential: Associate Data Scientist,
Kemendikbudristek, 2021.
• https://venturebeat.com/2020/08/18/how-duolingo-
uses-ai-in-every-part-of-its-app/

1
6
2. Machine Learning dan
Artificial Neural Network
A beginner’s guide to machine learning #2
Presented by
Ahmad Ataka Awwalur Rizqi, S.T., Ph.D.
Universitas Gadjah Mada, PIJAR Foundation
Outline Machine Learning dan ANN

• Review Singkat Machine Learning


• Klasifikasi Machine Learning dan Aplikasinya
• Artificial Neural Network, Back-propagation, dan Gradient Descent

1
8
Review
Machine Learning
“ A field of study that gives computers the
ability to learn without being explicitly
programmed.”

Attributed to Arthur Samuel

1
9
Klasifikasi Machine Learning

2
0
Supervised Learning

1. Given set of examples (input-output pairs), learns a function that maps


from input to output
2. Examples:
• Using the pixels of an image to detect the presence or absence of a cat
• Using the movies you’ve liked to predict more movies you may like
• Using someone’s words to predict whether they’re happy or sad
• Using weather sensor data to predict the probability of rain
3. Require labelled data.
4. Classified into regression and classification.
5. Most-used machine learning today!
2
1
Unsupervised
Learning
1. Given set of examples without label, detect potentially useful clusters of input
2. The dataset that it transforms into is not previously known or understood.
3. Unlike supervised learning, there is no “right answer” that you’re trying to get the
model to duplicate. You just tell an unsupervised algorithm to “find patterns in this data
and tell me about them.”

2
2
Reinforcement Learning
1. Agent learns from a series of reinforcements
(rewards or punishments)
2. The actions taken by the agent influence the
information that the environment provides to the
agent—that is, the agent receives direct feedback
on the actions it takes

2
3
What I can do with
Reinfocement Learning

2
4
Mini-Quiz
• Match the following task with the correct types of
machine learning:
• Mengelompokkan berbagai customer (S/U/R)
• Menebak hari terjadinya kemacetan berdasarkan
data jumlah kendaraan tahun lalu (S/U/R)

2
5
Human Neuron vs Artificial Neural Network

2
6
Artificial Neural Network (ANN)
1. Computational learning system that uses a network of functions to understand and translate a
data input of one form into a desired output
2. It’s largely used for machine learning in various fields.
3. It uses an optimizer to update and find the optimal values of weights which helps the model to
minimize the error i.e. difference between the actual and predicted values.

2
7
How ANN learns?

1. Feedforward — calculates the


predicted value.

2. Calculate Loss – different between the


target value and the predicted value

3. Backpropagation — updates the


weights to minimize loss score. It’s the
central algorithm of ANNs.

2
8
Deep network and Deep learning
1. Deep neural networks are networks that have many layers (hence the “deep”).
2. It induces the model to learn layered representations of input data.

2
9
Recap Machine Learning
dan ANN
1. Machine learning can be classified into:
• Supervised Learning
• Unsupervised Learning
• Reinforcement Learning
2. Artificial neural network is at the heart of
modern machine learning
3. It is able to learn from dataset to translate a
data input of one form into a desired output
4. A modern architecture of ANN is the deep
network where a lot of layers are used

3
0
References
1. Andrew Trask. 2019. Grokking Deep Learning (1st. ed.). Manning Publications Co., USA.
2. Jon Krohn, Grant Beyleveld, and Aglae Bassens. 2019. Deep Learning Illustrated: A Visual,
Interactive Guide to Artificial Intelligence (1st. ed.). Addison-Wesley Professional.
3. Deep Reinforcement Learning
4. Materi Microcredential: Associate Data Scientist, Kemendikbudristek, 2021.
5. https://towardsdatascience.com/how-things-work-artificial-neural-networks-6b1291c43d75

3
1
3. Using My First Machine
Learning Hands-on
A beginner’s guide to machine learning #3
Presented by
Ahmad Ataka Awwalur Rizqi, S.T., Ph.D.
Universitas Gadjah Mada, PIJAR Foundation
Outline
Using My First Machine Learning Hands-on
• Review Singkat ANN
• Apa yang kamu butuhkan
• Praktek 1: IRIS Classifier
• Praktek 2: Stockmarket Predictor

3
3
Review Singkat
ANN

1. Feedforward
Calculates the predicted value.
2. Calculate Loss
Different between the target value
and the predicted value
3. Backpropagation
Updates the weights to minimize loss
score. It’s the central algorithm of
ANNs.

3
4
Yuk kita coba menggunakan machine learning!

3
5
Berkenalan dengan Python
1. Apa itu Python?
• Bahasa pemrograman tingkat tinggi

2. Keunggulan Python:
• Keterbacaan, Python mudah dibaca dan dipahami
• Efisien, memiliki library yang lengkap sehingga
penulisan coding dapat lebih sederhana
• Multifungsi, dengan menggunakan Python dapat
membuat website, aplikasi bidang robotika, aplikasi
bidang kecerdasan artifisial.
• Interoperabilitas, Python mampu berinteraksi dengan
Bahasa pemrograman lainnya
• Dukungan komunitas, Python merupakan program
open source, dan komunitas Python sangat aktif dalam
melakukan pengembangan

3
6
Python development
environment
1. Anaconda Distribution
(https://www.anaconda.com/distribution/)
2. Miniconda
(https://docs.conda.io/en/latest/miniconda.html)
3. Jupyter Notebook (https://jupyter.org/)
4. Python installer
(https://www.python.org/downloads/).
5. Google Colaboratory
(https://colab.research.google.com/).
6. Notebooks Azure (https://notebooks.azure.com/)
7. Github (https://github.com/)

3
7
Praktek 1: IRIS Classifier
• Go to: Google Colab http://ugm.id/handsonmspijar1

3
8
Praktek 2: Forecasting
• Go to: Google Colab http://ugm.id/handsonmspijar2

3
9
Recap Hands-on
• Machine learning bisa digunakan untuk
berbagai keperluan di berbagai sector
seperti klasifikasi hingga forecasting

• Python adalah salah satu Bahasa


pemrograman yang banyak menyediakan
library untuk membuat model machine
learning

• Machine learning, khususnya supervised


learning, butuh dataset yang cukup agar bisa
bekerja dengan baik. Maka, ketersediaan
data yang berkualitas menjadi krusial.

4
0
4. The future of Machine Learning
in the world and Indonesia
A beginner’s guide to machine learning #4
Presented by
Ahmad Ataka Awwalur Rizqi, S.T., Ph.D.
Universitas Gadjah Mada, PIJAR Foundation
Outline
The future of Machine Learning in the world
and Indonesia
• Recap konsep-konsep penting
• Perkembangan AI di dunia dan Indonesia
• Future of AI and Machine Learning
• Closing Statement

4
2
What we have learned?

• Apa itu pembelajaran mesin?


• Apa jenis-jenis pembelajaran mesin?
• Apa itu artificial neural network?
• Sebutkan contoh-contoh penggunaan
machine learning

4
3
Perkembangan AI di dunia

4
4
4
5
4
6
Future of AI

4
7
Closing Statement

• AI dan machine learning adalah salah satu


teknologi penting yang bisa mengoptimalkan
berbagai sector pekerjaan

• Menguasai machine learning dan


menggunakannya dalam konteks Indonesia
menjadi krusial di era industry 4.0

• Let’s build our country with technology!

4
8
Terima
Kasih.
CONFIDENTIAL

4
9

You might also like