Outline of AIPT. Renom
InfrastrukturIndonesia
2
Agenda
Opening - introduction of speaker and company -
Introduction - purpose of seminar –
Outline of AI and data scientist
Introduction of GUI framework to play with data (ReNom TDA/TAG/IMG)
Introduction of data science bootcamp; make.ai
Closing
3
Introduction: About Me – Kenalkan, nama saya Takamine TAKINO. -
2013 – 2014: Research Assistant for ESDM/JICA in Jakarta
2015: Graduated from MS. of Computer Science
<Thesis: Intelligent Lighting System Using Illuminance and Luminosity Database>
2015 – 2016: System engineer for power system in Hitachi Inc.
2016 – Present: Senior data scientist & AI business consultant for GRID Inc., and R&D advisor for PT. RII.
2017 – Present: Founder of AI Indonesia (510 members community)
Personal History
Hobby
Travel: udah ke … Bali, Surabaya, Bromo, Bandung, Yogya, Solo, Bogor, Pulau Seribu, Komodo, Manado, Kalimantan
Watching movie: udah nonton … Ayat ayat cinta (give me recommendation of Indonesian movies)
CONFIDENTIAL この文書は、著作権法及び不正競争防止法上の保護を受けております。文書の一部あるいはすべてについて、株式会社グリッドの許諾を得ずに、いかなる方法においても無断で複写、複製、転記、転載を行うことは禁じられています。
Introduction: About Us
2015 – now
AI/IoT businessUser-friendly machine learning framework
named ReNom & Bigdata analysis service
2009 – now
Renewable Energy businessPower plant development & management
Representative Director Hideki Nakamura
Headquarter 6F Ao building building, Kitaaoyama3-11-7, Minato-ku, Tokyo
Indonesia Representative Office Citywalk, Jalan KH. Mas Mansyur Kav.121
5
Purpose of Seminar
We would like you to aware the impact of Artificial Intelligence
We would like you to get valuable job
We would like to lead you to do that
6
Agenda
Opening - introduction of speaker and company -
Introduction - purpose of seminar –
Outline of AI and data scientist
Introduction of GUI framework to play with data (ReNom TDA/TAG/IMG)
Introduction of data science bootcamp; make.ai
Closing
7
Image / Security & Medical
7
8
Recognition of Handwritten Character
8
9
Self Driving Car
10
Image - Amazon Echo Look
10
11
Image – Draw the picture
11
12
Language – Google Translate
12
13
Language / Smart Speaker
13
Microsoft/Cortana
Amazon/Alexa
Google/Home
SONOS/PLAY:1
Apple/Siri
1414
“AI would be the ultimate
version of Google” Oct, 2000
“We will move from mobile first to an AI first” Apr, 2016
15
Classify animals
15
16
Weight(kg)
Height(cm)
giraffe
hippo
zebra
Abstract character
16
17
Weight(kg)
Height(cm)
hippo
zebra
giraffe
Draw border
17
18
Weight(kg)
Height(cm)
hippo
zebra
giraffe
New data
18
19
Weight Color SpeedHeight Input data
zebra giraffe Answer
Logic
19
Neural Network
20
500kg W/B 60km/h2m Input data
zebra giraffe Answer
特徴判断ロジック
Neural Network
20
21
700kg Yellow 40km/h4m Input data
zebra giraffe Answer
特徴判断ロジック
Neural Network
21
22
Female Male
Picture Image
22
Female Male
Picture Image
23
Learn with Bigdata
23
Apple Apple Apple Apple Apple Apple Apple
24
1. Abstract character
2. Draw border
AI is Simple
24
25
Inputlayer Hidden
layer
Outputlayer
Inputlayer Hidden
layerHiddenlayer
Outputlayer
3Layers 4Layers
Machine Learning vs Deep Learning
25
26
Agenda
Opening - introduction of speaker and company -
Introduction - purpose of seminar –
Outline of AI and data scientist
Introduction of GUI framework to play with data (ReNom TDA/TAG/IMG)
Introduction of data science bootcamp; make.ai
Closing
27
Daily Life of “Data Scientist”
Visualize & explore data
Discuss the analysis
approach
Preprocessing
Implementmachine learning
algorithm
Test the accuracy
Optimize parameter/
change model
Report resultto project
manager/client
28
Data Visualization
29
Data Preprocessing
30
Impalement Machine Learning Algorithm
31
Elements of Data Scientist – Everyone says –
Python, R, SQL
Retail, Finance, Energy, Food, Logistics, Agriculture, Automobile, Fashion etc
Linear algebra, probability,linear regression, SVM, dimensional reduction etc
32
Elements of Data Scientist – But actual work is –
Python, R, SQLLinear algebra, probability,linear regression, SVM, dimensional reduction etc
Business knowledge
Retail, Finance, Energy, Food, Logistics, Agriculture, Automobile, Fashion etc
33
Carrier of Data Scientist
Do you have strong passion in certain industry?
Work in the company
Join data science bootcamp
Are you self-learner?
E-learning
Be data science intern
OJT/Projects
Be data scientist
Join online/offline community too
E-commerce for fashion, energy, retail, bank etc
Job related to data
Others
Yes No
Yes No Yes No
34
Agenda
Opening - introduction of speaker and company -
Introduction - purpose of seminar –
Outline of AI and data scientist
Introduction of GUI framework to play with data (ReNom TDA/TAG/IMG)
Introduction of data science bootcamp; make.ai
Closing
35
GUI Framework for Computer Vision –ReNom TAG -
36
GUI Framework for Computer Vision –ReNom IMG -
37
GUI Framework for Computer Vision –ReNom IMG -
38
Why they survive / died? – Power of ReNom TDA -
382018/2/23
Survivor (Blue:Dead, Red:Survive)
Gender (Blue:Female, Red:Male)
Fare (Blue:Low, Red:High)
Room Grade (Blue:High, Red:Low)
Age (Blue:Low, Re:High)
Family (Blue:0, Red:4)
39
Agenda
Opening - introduction of speaker and company -
Introduction - purpose of seminar –
Outline of AI and data scientist
Introduction of GUI framework to play with data (ReNom TDA/TAG/IMG)
Introduction of data science bootcamp; make.ai
Closing
40
KURIKULUMEDUKASI
Day Material End Goal
1 Python 1 Peserta dapat menggunakan bahasa pemrograman Python2 Python 2
3 SQL 1SQL untuk mengoperasikan database
4 SQL 2 & Python Integration
41
KURIKULUMEDUKASI
Day Material End Goal
5 Basic Stat & MathPeserta dapat memahami statistika dan matematika dasar untuk kebutuhan analisis data
6Kaggle & Git Tutorial Peserta dapat memahami penggunaan & benefit
dari platform Kaggle & Git
7 Visualization Peserta dapat mengekstrak informasi, memvisualisasi kesimpulan dan membersihkan data mentah8 Preprocessing
42
KURIKULUMEDUKASI
Day Material End Goal
9 Linear & Logistic Reg. Peserta dapat membuat model Art i f ic ia lIntelligence yang menghasilkan prediksi/deteksil a l u m e m a h a m i c a r a k e r j a n y a10 SVM
11 Random Forest Peserta dapat membuat model Artificial Intelligence yang menghasilkan prediksi/deteksi lalu memahami cara kerjanya12 Clustering
43
KURIKULUMEDUKASI
Day Material End Goal
13 Ensemble Peserta dapat membuat model Art i f ic ia lIntelligence yang menghasilkan prediksi/deteksil a l u m e m a h a m i c a r a k e r j a n y a14
Perceptron & FCNN
15Dimensionality Reduction Peserta dapat menggunakan metode - metode
validasi & optimisasi untuk meningkatkan hasil model Artificial Intelligence yang telah dibuat16
Validation & Optimization
44
KURIKULUMEDUKASI
Day Material End Goal
17 ReportingPeserta dapat membuat laporan akhir dari kasus yang telah dipecahkan dengan menggunakan metode Data Science
18 Study Case 1Peserta dapat menerapkan metode - metode data science ke dalam studi kasus 1 dan menyelesaikannya
19 Study Case 1Peserta dapat menerapkan metode - metode data science ke dalam studi kasus 1 dan menyelesaikannya
45
KURIKULUMEDUKASI
Day Material End Goal
20 Study Case 2Peserta dapat menerapkan metode - metode data science ke dalam studi kasus 2 dan menyelesaikannya
21 Study Case 2Peserta dapat menerapkan metode - metode data science ke dalam studi kasus 2 dan menyelesaikannya
22 Study Case 3Peserta dapat menerapkan metode - metode data science ke dalam studi kasus 3 dan menyelesaikannya
23 Study Case 3Peserta dapat menerapkan metode - metode data science ke dalam studi kasus 3 dan menyelesaikannya
24 Final Test Data SciencePengukuran hasil akhir peserta setelah menyelesaikan pembelajaran Data Science
GROUPMENTOR
Head of Data ResearcherAndri Danusasmita
Head of Data EngineerRey Steven Octoviano
Data ScientistCitra Hasana Sagala
Senior Data ScientistLoya Jirga
Technical AdvisorTakamine Takino
Data ScientistDhanang Hadhi Sasmita