際際滷

際際滷Share a Scribd company logo
磯Μ企殊磯
Since 2003
2019  01-08
蟲覩朱蟲 觜一危 覿  D&A
Ai&ML Seminar in N3N cloud
INDEX
1. Ai vs ML vs DL
2. Ai Researcher & Data Scientist 蟆
GPU螳  伎
3. 一危 覿 Process 覦 蟯 Tool 螳
4. 蠍郁 Demo
AI vs ML vs DL
1
蠍郁(ML), ル(DL)
瑚概讌(AI) 蟲 蠍一 譴 企
Deep Learning 
Network capable
of adapting itself
to new data
Artificial Intelligence  Computers
with the ability to reason as humans
Machine Learning 
Computers with the ability to
learn without being explicitly
programed
AI > ML > DL
1. AI vs ML vs DL
瑚概讌 :
蠍郁襦 覿 襷れ伎 讌ワ
1. AI vs ML vs DL
1. AI vs ML vs DL
AI
讌蠍磯 覦
蠍郁 
1. AI vs ML vs DL
 蟲覃 2螳願 譴螳 覿覿 讚覃,
襷  螳 リ啓 覈企朱 8企. 
AI
讌蠍磯 覦
蠍郁 
讌蠍磯 覦
讌襷, 讌  覯企  覦
 殊殊 讌蠍磯朱 蠏豺 襷 蟆 覿螳
瑚 覲螳  ル伎 所 語螻 牛.
讌蠍磯 覦
1. AI vs ML vs DL
 蟆り 轟 瑚概讌リ骸 語螻狩 覿殊 讌 讌レ 殊
 覓 伎 襾語, 糾骸 蟯 譯殊 蟇一 蟯 覲伎伎 .
蟆郁骸朱 蠍郁  郁規れ 朱 覦 企れ 蟆朱
朱朱 麹  覦讌 覈詩り 蠎. [Langely2011] 
讌蠍磯 覦
1. AI vs ML vs DL
讌蠍磯 覦 螻
1. AI vs ML vs DL
AI
讌蠍磯 覦
蠍郁 
蠍郁 
1. AI vs ML vs DL
 :
蟆渚 蟆郁骸襦 , 觜蟲 讌  覲
蠏 レ 覲.  讌 給 螻殊 [蟲襴所記伎2017]
蠍郁 
1. AI vs ML vs DL
vs
れ 譴 螻企?
蠍郁 
1. AI vs ML vs DL
vs
れ 譴 豢?
蠍郁 
1. AI vs ML vs DL
(Learning)企?
れ 譴 覦煙襴?
蠍郁 
1. AI vs ML vs DL
蠍郁
企 貉危 襦蠏碁 T朱  .
 襦蠏碁 焔レ P朱 豌襦 螳 
蟆渚 E襯 牛 焔レ 螳る
 襦蠏碁 旧 り 襷  .
[Mitchell1997]
蠍郁 
1. AI vs ML vs DL
蠍郁汲
襦 一危, 讀 螻手碓 蟆渚 伎
焔 蠍一 豕襦 襦蠏碁覦 
[Alpaydin2010]
螳讌  企語襯 貉危郁 牛
襦 煙ロ 讌 覦煙襴 語 覿襯 襦蠏碁覦 
蠍郁 
1. AI vs ML vs DL
7 + 8 = 15
 豢レ襴 螻殊
蠍一ヾ 襦蠏碁
() (豌襴螻殊) (豢)
襦蠏碁襾瑚 伎  蟆
蟲螻  蟆
蠍郁 
1. AI vs ML vs DL
襦蠏碁襾瑚 伎  蟆
蠍郁 
1. AI vs ML vs DL
7 + 8 = 15
 豢レ襴 螻殊
蠍郁 
() (豌襴螻殊) (豢)
蟲螻  蟆
願 覓企
蠍郁 
1. AI vs ML vs DL
蠍郁 
1. AI vs ML vs DL
願 覓願
覓
蠍郁: 覓語 旧 牛 觚覦るゼ 襷 蟆
蟲蟆 給 觚覦る 豢 襦 覓語襯  覦  襷  豢ロ
蠍郁 
1. AI vs ML vs DL
ル (Deep Learning) :
蠍郁旧  覿朱 一危磯ゼ 貉危郁 豌襴 螳ロ
 覯″磯 蠏碁 煙朱 螻
企ゼ 牛 覈語 蟲豢 郁規襯 も
譟一磯Μ 企殊磯
ル
1. AI vs ML vs DL
DATA
 一危
觜 一危
ex)   一危
ex) ろ, 企語, ,  
蠍郁 旧  一危
1. AI vs ML vs DL
(X) 豢(Y)
1 2
2 4
3 6
4 8
5 10
10 20
100 200
 一危一 蠍郁
1. AI vs ML vs DL
(X)
1
2
3
4
5
10
100
Y = 2X
豢(Y)
2
4
6
8
10
20
200
 一危一 蠍郁
1. AI vs ML vs DL
 (X1)  襯觜(X2)  蠏 豢 (X3) (X4)  譟語 覿(X5) 焔(Y)
2,000,000 200,000 0 22 1 
1,000,000 100,000 100,000 20 0 
50,000,000 - 40,000,000 45 1 
3,000,000 1,000,000 2,000,000 30 1 
4,000,000 300,000 1,000,000 30 0 
5,000,000 450,000 3,000,000 34 0 
150,000,000 1,000,000 10,000,000 50 0 
 一危一 蠍郁
1. AI vs ML vs DL
DATA
 一危
觜 一危
ex)   一危
ex) ろ, 企語, ,  
蠍郁 旧 一危
1. AI vs ML vs DL
KB蟲覩殊拘  08/17
17:00 20,000 譯殊 企
 
*襴* 牛 襴伎 
語襴蟆給. る 蠍郁
 危 覩語朱 
(蟯螻) 豌企
蠑語 
11 企
觜 覿x

[蟲覲企瑚] <語 企
るゼ 螻褐 るジ> 豈 
殊 
[企ろ襴]
覃る 
50% ~ 60%

 讌碁, 轟 る 覲朱? 
rrr 
. 豌 豢
20% 覓伎 
觜 一危一 蠍郁_ろ
1. AI vs ML vs DL
螻
襷豺
覦煙襴
觜 一危一 蠍郁_企語
1. AI vs ML vs DL
Deep Learning : DNN(Deep Neural Network)
ル
1. AI vs ML vs DL
Deep
Learning
2
Deep
Learning
1
ル
1. AI vs ML vs DL
1. 誤磯 覿  一危郁 覿伎
2. 螳 GPU 煙
3. 襦 企
(CNN, 蠏碁誤 覃 覓語 願屋, れ 蠏 蠍磯 煙, 豸給 觜 蠍磯 螳覦 )
ル 蠍一  
1. AI vs ML vs DL
Ai Researcher & Data Scientist 蟆
GPU螳  伎
2
vs
CPU (譴豌襴レ) GPU (蠏碁 豌襴 レ)
2. Ai Researcher & Data Scientist蟆
GPU螳  伎
 蠍郁旧 GPU 螳  伎
2. Ai Researcher & Data Scientist蟆
GPU螳  伎
 蠍郁旧 GPU 螳  伎
CPU GPU
2. Ai Researcher & Data Scientist蟆
GPU螳  伎
 蠍郁旧 GPU 螳  伎
2. Ai Researcher & Data Scientist蟆
GPU螳  伎
 蠍郁旧 GPU 螳  伎
2. Ai Researcher & Data Scientist蟆
GPU螳  伎
 蠍郁旧 GPU 螳  伎
2. Ai Researcher & Data Scientist蟆
GPU螳  伎
れ 蟲覩朱蟲 螻給 覯 UI
2. Ai Researcher & Data Scientist蟆
GPU螳  伎
れ 蟲覩朱蟲 螻給 覯 UI
2. Ai Researcher & Data Scientist蟆
GPU螳  伎
れ 蟲覩朱蟲 螻給 覯 UI
2. Ai Researcher & Data Scientist蟆
GPU螳  伎
れ 蟲覩朱蟲 螻給 覯 UI
蠍郁旧 譯朱  Tool &
一危 覿 谿
3
3. 蠍郁旧 譯朱  Tool & 一危 覿 谿
蠍郁旧 譯朱  Tool
3. 蠍郁旧 譯朱  Tool & 一危 覿 谿
Deep Learning Framework
3. 蠍郁旧 譯朱  Tool & 一危 覿 谿
Deep Learning Framework
3. 蠍郁旧 譯朱  Tool & 一危 覿 谿
Ai&ML Seminar in N3N cloud
一危
讌
覓語


一危
覿
一危
豌襴
一危
覈碁

蟆讀
豕譬 覈
ろ
1 2 3 4 5 6 7
 企 

蟆瑚?
 Download
 API
 Crawling
 覿
 蟯蟯螻
 . . .
 Missing Value
 Select
Require Data
 Cleansing
 螻襴讀

 覈 
 覈 蟆讀
 豕譬 覈
ろ
80% 20%
3. 蠍郁旧 譯朱  Tool & 一危 覿 谿
蠍郁旧 谿
蠍郁 Demo
4
螳矧

More Related Content

Similar to Ai&ML Seminar in N3N cloud (20)

Spark & Zeppelin 襾語 れ 蠍
Spark & Zeppelin  襾語 れ 蠍Spark & Zeppelin  襾語 れ 蠍
Spark & Zeppelin 襾語 れ 蠍
Taejun Kim
Ai(瑚概讌) & ML(襾語) 101 Part1
Ai(瑚概讌) & ML(襾語) 101 Part1Ai(瑚概讌) & ML(襾語) 101 Part1
Ai(瑚概讌) & ML(襾語) 101 Part1
Donghan Kim
貉MDS_DataRobot Overview
貉MDS_DataRobot Overview貉MDS_DataRobot Overview
貉MDS_DataRobot Overview
HANCOM MDS
[WeFocus] 瑚概讌_ル_麹 覲 _蟾燕_201902_v1
[WeFocus] 瑚概讌_ル_麹 覲 _蟾燕_201902_v1[WeFocus] 瑚概讌_ル_麹 覲 _蟾燕_201902_v1
[WeFocus] 瑚概讌_ル_麹 覲 _蟾燕_201902_v1
Luke Sunghyun Kim
AI_introduction and requirements(2024.05.12).pdf
AI_introduction and requirements(2024.05.12).pdfAI_introduction and requirements(2024.05.12).pdf
AI_introduction and requirements(2024.05.12).pdf
Lee Chanwoo
4 蟲IBM螻 蟷 螻給 ろ 誤 碁碁-AI
4 蟲IBM螻 蟷 螻給 ろ 誤 碁碁-AI4 蟲IBM螻 蟷 螻給 ろ 誤 碁碁-AI
4 蟲IBM螻 蟷 螻給 ろ 誤 碁碁-AI
Tommy Lee
ル 瑚 覓誤蠍 覦 覿
ル 瑚 覓誤蠍 覦 覿ル 瑚 覓誤蠍 覦 覿
ル 瑚 覓誤蠍 覦 覿
Ubuntu Korea Community
襾語 覦 ル 蠍一 (2020.01.07)
襾語 覦 ル 蠍一 (2020.01.07)襾語 覦 ル 蠍一 (2020.01.07)
襾語 覦 ル 蠍一 (2020.01.07)
KYOYOON JUNG
過メ 求求 求求 AI =戟求=AI ≡=釈=,
過メ 求求 求求 AI =戟求=AI ≡=釈=, 過メ 求求 求求 AI =戟求=AI ≡=釈=,
過メ 求求 求求 AI =戟求=AI ≡=釈=,
Daehee Han
襷語 Big Data Recommendation Products - 伎 一危磯ゼ 牛 伎殊 豸″
襷語 Big Data Recommendation Products - 伎 一危磯ゼ 牛 伎殊 豸″襷語 Big Data Recommendation Products - 伎 一危磯ゼ 牛 伎殊 豸″
襷語 Big Data Recommendation Products - 伎 一危磯ゼ 牛 伎殊 豸″
Evion Kim
Introduction to industrial engineering
Introduction to industrial engineeringIntroduction to industrial engineering
Introduction to industrial engineering
Kang Pilsung
[2A7]Linkedin'sDataScienceWhyIsItScience
[2A7]Linkedin'sDataScienceWhyIsItScience[2A7]Linkedin'sDataScienceWhyIsItScience
[2A7]Linkedin'sDataScienceWhyIsItScience
NAVER D2
求求 ≡ 求 =求メ求梶梶 =
求求 ≡ 求 =求メ求梶梶 =求求 ≡ 求 =求メ求梶梶 =
求求 ≡ 求 =求メ求梶梶 =
Jaimie Kwon (蟠覈)
襾語 螳 觸螳蠍
襾語 螳 觸螳蠍襾語 螳 觸螳蠍
襾語 螳 觸螳蠍
Heejae Jeong
求メ求 求求 求=求 1510600 ≡過メ 求≡ 求 求 =求求
求メ求 求求 求=求 1510600 ≡過メ   求≡ 求 求 =求求求メ求 求求 求=求 1510600 ≡過メ   求≡ 求 求 =求求
求メ求 求求 求=求 1510600 ≡過メ 求≡ 求 求 =求求
merry818
AI蟆 伎襯 覓視: る 螳ロ 瑚概讌(XAI: eXplainable AI)
AI蟆 伎襯 覓視: る 螳ロ 瑚概讌(XAI: eXplainable AI)AI蟆 伎襯 覓視: る 螳ロ 瑚概讌(XAI: eXplainable AI)
AI蟆 伎襯 覓視: る 螳ロ 瑚概讌(XAI: eXplainable AI)
Hansung University
T ka kr_4th
T ka kr_4thT ka kr_4th
T ka kr_4th
Subin An
Coding interview
Coding interviewCoding interview
Coding interview
Soohan Ahn
Spiral 覈 蠍磯 る AI 蟲.pdf
Spiral 覈 蠍磯 る AI 蟲.pdfSpiral 覈 蠍磯 る AI 蟲.pdf
Spiral 覈 蠍磯 る AI 蟲.pdf
MyungHoKim10
[D2 CAMPUS] Tech meet-up `data science` 覦襭
[D2 CAMPUS] Tech meet-up `data science` 覦襭[D2 CAMPUS] Tech meet-up `data science` 覦襭
[D2 CAMPUS] Tech meet-up `data science` 覦襭
NAVER D2
Spark & Zeppelin 襾語 れ 蠍
Spark & Zeppelin  襾語 れ 蠍Spark & Zeppelin  襾語 れ 蠍
Spark & Zeppelin 襾語 れ 蠍
Taejun Kim
Ai(瑚概讌) & ML(襾語) 101 Part1
Ai(瑚概讌) & ML(襾語) 101 Part1Ai(瑚概讌) & ML(襾語) 101 Part1
Ai(瑚概讌) & ML(襾語) 101 Part1
Donghan Kim
貉MDS_DataRobot Overview
貉MDS_DataRobot Overview貉MDS_DataRobot Overview
貉MDS_DataRobot Overview
HANCOM MDS
[WeFocus] 瑚概讌_ル_麹 覲 _蟾燕_201902_v1
[WeFocus] 瑚概讌_ル_麹 覲 _蟾燕_201902_v1[WeFocus] 瑚概讌_ル_麹 覲 _蟾燕_201902_v1
[WeFocus] 瑚概讌_ル_麹 覲 _蟾燕_201902_v1
Luke Sunghyun Kim
AI_introduction and requirements(2024.05.12).pdf
AI_introduction and requirements(2024.05.12).pdfAI_introduction and requirements(2024.05.12).pdf
AI_introduction and requirements(2024.05.12).pdf
Lee Chanwoo
4 蟲IBM螻 蟷 螻給 ろ 誤 碁碁-AI
4 蟲IBM螻 蟷 螻給 ろ 誤 碁碁-AI4 蟲IBM螻 蟷 螻給 ろ 誤 碁碁-AI
4 蟲IBM螻 蟷 螻給 ろ 誤 碁碁-AI
Tommy Lee
襾語 覦 ル 蠍一 (2020.01.07)
襾語 覦 ル 蠍一 (2020.01.07)襾語 覦 ル 蠍一 (2020.01.07)
襾語 覦 ル 蠍一 (2020.01.07)
KYOYOON JUNG
過メ 求求 求求 AI =戟求=AI ≡=釈=,
過メ 求求 求求 AI =戟求=AI ≡=釈=, 過メ 求求 求求 AI =戟求=AI ≡=釈=,
過メ 求求 求求 AI =戟求=AI ≡=釈=,
Daehee Han
襷語 Big Data Recommendation Products - 伎 一危磯ゼ 牛 伎殊 豸″
襷語 Big Data Recommendation Products - 伎 一危磯ゼ 牛 伎殊 豸″襷語 Big Data Recommendation Products - 伎 一危磯ゼ 牛 伎殊 豸″
襷語 Big Data Recommendation Products - 伎 一危磯ゼ 牛 伎殊 豸″
Evion Kim
Introduction to industrial engineering
Introduction to industrial engineeringIntroduction to industrial engineering
Introduction to industrial engineering
Kang Pilsung
[2A7]Linkedin'sDataScienceWhyIsItScience
[2A7]Linkedin'sDataScienceWhyIsItScience[2A7]Linkedin'sDataScienceWhyIsItScience
[2A7]Linkedin'sDataScienceWhyIsItScience
NAVER D2
求求 ≡ 求 =求メ求梶梶 =
求求 ≡ 求 =求メ求梶梶 =求求 ≡ 求 =求メ求梶梶 =
求求 ≡ 求 =求メ求梶梶 =
Jaimie Kwon (蟠覈)
襾語 螳 觸螳蠍
襾語 螳 觸螳蠍襾語 螳 觸螳蠍
襾語 螳 觸螳蠍
Heejae Jeong
求メ求 求求 求=求 1510600 ≡過メ 求≡ 求 求 =求求
求メ求 求求 求=求 1510600 ≡過メ   求≡ 求 求 =求求求メ求 求求 求=求 1510600 ≡過メ   求≡ 求 求 =求求
求メ求 求求 求=求 1510600 ≡過メ 求≡ 求 求 =求求
merry818
AI蟆 伎襯 覓視: る 螳ロ 瑚概讌(XAI: eXplainable AI)
AI蟆 伎襯 覓視: る 螳ロ 瑚概讌(XAI: eXplainable AI)AI蟆 伎襯 覓視: る 螳ロ 瑚概讌(XAI: eXplainable AI)
AI蟆 伎襯 覓視: る 螳ロ 瑚概讌(XAI: eXplainable AI)
Hansung University
T ka kr_4th
T ka kr_4thT ka kr_4th
T ka kr_4th
Subin An
Coding interview
Coding interviewCoding interview
Coding interview
Soohan Ahn
Spiral 覈 蠍磯 る AI 蟲.pdf
Spiral 覈 蠍磯 る AI 蟲.pdfSpiral 覈 蠍磯 る AI 蟲.pdf
Spiral 覈 蠍磯 る AI 蟲.pdf
MyungHoKim10
[D2 CAMPUS] Tech meet-up `data science` 覦襭
[D2 CAMPUS] Tech meet-up `data science` 覦襭[D2 CAMPUS] Tech meet-up `data science` 覦襭
[D2 CAMPUS] Tech meet-up `data science` 覦襭
NAVER D2

More from HWANGTAEYONG (6)

DNN-Based Prediction Model for Spatial-Temporal Data
DNN-Based Prediction Model for Spatial-Temporal DataDNN-Based Prediction Model for Spatial-Temporal Data
DNN-Based Prediction Model for Spatial-Temporal Data
HWANGTAEYONG
Prediction of bus riders at the time of departure
Prediction of bus riders at the time of departurePrediction of bus riders at the time of departure
Prediction of bus riders at the time of departure
HWANGTAEYONG
Strategies to cluster customers using machine learning and send differentiate...
Strategies to cluster customers using machine learning and send differentiate...Strategies to cluster customers using machine learning and send differentiate...
Strategies to cluster customers using machine learning and send differentiate...
HWANGTAEYONG
PBL : How to live long and healthy in Health Promotion House
PBL : How to live long and healthy in Health Promotion HousePBL : How to live long and healthy in Health Promotion House
PBL : How to live long and healthy in Health Promotion House
HWANGTAEYONG
Development of a forecasting model for expected profit for lineage users
Development of a forecasting model for expected profit for lineage usersDevelopment of a forecasting model for expected profit for lineage users
Development of a forecasting model for expected profit for lineage users
HWANGTAEYONG
伎 襦る
伎   襦る伎   襦る
伎 襦る
HWANGTAEYONG
DNN-Based Prediction Model for Spatial-Temporal Data
DNN-Based Prediction Model for Spatial-Temporal DataDNN-Based Prediction Model for Spatial-Temporal Data
DNN-Based Prediction Model for Spatial-Temporal Data
HWANGTAEYONG
Prediction of bus riders at the time of departure
Prediction of bus riders at the time of departurePrediction of bus riders at the time of departure
Prediction of bus riders at the time of departure
HWANGTAEYONG
Strategies to cluster customers using machine learning and send differentiate...
Strategies to cluster customers using machine learning and send differentiate...Strategies to cluster customers using machine learning and send differentiate...
Strategies to cluster customers using machine learning and send differentiate...
HWANGTAEYONG
PBL : How to live long and healthy in Health Promotion House
PBL : How to live long and healthy in Health Promotion HousePBL : How to live long and healthy in Health Promotion House
PBL : How to live long and healthy in Health Promotion House
HWANGTAEYONG
Development of a forecasting model for expected profit for lineage users
Development of a forecasting model for expected profit for lineage usersDevelopment of a forecasting model for expected profit for lineage users
Development of a forecasting model for expected profit for lineage users
HWANGTAEYONG

Ai&ML Seminar in N3N cloud