際際滷

際際滷Share a Scribd company logo
TENSORFLOW 螳
2016.12.02 (蠍) 危
Index
 ル螻 襦
 襦 れ
 蠏覿
 覲螳 蟯螻  覈
 觜 蟆曙 螳覯 螻襴讀
 螻襴讀 ろ
ル螻 襦
ル螻 襦
ル
 蟆暑 螻襴讀 譯朱 

 殊碁 螻襴讀
 CNN螻 RNN
 CNN(炎咳 蟆暑)
 企語 伎 覓殊牡 語
 RNN( 蟆暑)
 覯螻 煙語
襦
 豕蠏殊 ル 郁規 
襴  殊企襴
襦
 蠍磯蓋 蟲譟
 一危 襦 蠏碁襯  豺 一一 .
 蠏碁 碁 :  一
 蠏碁 讌 : れ姶 一危 覦一
 豺 一一 蠍壱碁  蠏碁 蟲譟磯ゼ 襷り 豌襴.
 レ
 CPU, GPU レ 覈 伎 螳
 覈覦 , 64觜 襴れ 覈 伎 螳
 覲企 覈
 襷 覲企ゼ 覈磯螻 ろ 伎.
 谿瑚 : https://goo.gl/6Cyr20
襦 vs 襦 觜
襦
 ル 一危磯ゼ 螳讌螻
襾語 螻襴讀 襷
れ 覈語 貅譴
.
襦 觜
  覈語 企殊伎
碁覿 れ伎る 
一危磯ゼ 覦伎.
襦 れ
#0 伎螻 伎宛 れ
 伎 れ
 伎 螳覦 蟆曙 伎宛 れ
#1 docker れ 蟆 れ
 襦 襴れ 覩襦 螳 襾語   譯朱
docker tool box襯 れ
 docker quick start terminal ろ
 docker-machine create vdocker d virtualbox
 豕豐 1 ろ!
 docker-machine ls
 vdocker 螳 Running語 誤
 Stopped企朱 docker-machine start vdocker
 磯碁 譬襭
#1 docker れ 蟆 れ
#2 cmd docker 企語 れ企
 cmd ろ
  覈  (伎郁鍵 譯殊.)
 FOR /f "tokens=*" %i IN ('docker-machine env --shell
cmd default') DO %i
 docker 企語 れ企 (豕豐 1 ろ!)
 docker run v /c/Users/ML:/home/ML it python bash
  伎宛 襦 蟆暑襯 襦  譯手鍵
 焔 企語覈 誤蠍
  : docker ps a
  : docker ps rm (image name)
 る docker stop (image name) 螻  覈
#3 焔 貉企 
  覦 
 docker restart (image name)
 docker attach (image name) + Enter (所 覯蠏!)
 覃 docker start (image name)   attach ろ
#4 襦 れ
 https://www.tensorflow.org/versions/r0.11/get_started/os_set
up.html
  貉危 蟆曙 襷 覈轟  襦 れ
 襭覃  襦 启襦 企 襦碁ゼ
ろ企蓋!
 cd (path)
 python (file name)
 朱  覿覿蟾讌 ロ螻 Tabるゼ 襯企 企 朱
#5 れ 襭  れ ロ 螻殊
#6 螳  ろ
: 螻襴讀 蠍一 > 語  > 一 ろ!
 placeholder
 襦蠏碁 ろ 譴 螳 覲蟆渚 
 覯襴 覲
 tensor
  蠍 螳 れ姶 一危 覦

 mul()
  譟一  襦螳 
螻牛 襷  一  譴

 Session()
 語 燕朱 襦蠏碁
襦 殊企襴 語
Note that
 襦 一郁骸 一危一  覈 覲企 蠏碁
蟲譟  ル
 碁 :  一
 一危 リ骸 豢レ 豺襯 蟇磯 ル 覲襯 所 
 讌 :  螳螻 豢 螳朱 郁屋 碁 螳 蟯螻襯  + 
襦 蠍磯蓋 襭蟲譟一  企!
 蠏 覿
覲 螳 蟯螻  覈
觜  蟆曙 螳覯 螻襴讀
螻襴讀 ろ
 蠏 覿(linear regression)企?
 覲 伎 蟯螻襯 覿   糾 覦
覯
 襴暑 xi,  b 譬覲 y螳 蟯螻襯 覈碁
 蠏 れ蠏
 蠏 :  覲 螳 蟯螻 蟆曙
 れ蠏 :  螳 覲襯 る 蟆曙
覲 螳 蟯螻  覈 : y =
W*x+b
 伎姶 譬螻 一危磯ゼ 燕 伎 襦蠏碁 

   一危一 襷 豕 讌 谿城 覓語
觜 (cost function) 蟆曙 螳覯
螻襴讀(gradient descent)
 覓語 :  一危磯覿 豢レ 豢    
螻襴讀 り鍵
 W b  螳 襷り覲襦 覈  螳
  一危 x_data襯 伎 豢 一危 y_data襯 襷  
豕 襷り覲 W b襯 谿剰鍵
觜  蟆曙 螳覯 螻襴讀
 觜(=れ姶)
 覦覲旧 殊企 襷 螳 覿襯 誤蠍  朱 譬
() 讌語 豸′
 W b襯 襷り覲襦 覦 讌 朱 一危一  襷讌襯
蠍一朱 れ姶 螳 襴
 蠍一 觜襦 蠏螻煙れ姶(mean square error) 
 れ 螳螻 螻襴讀 覦覲給 豢 螳 伎 蟇磯Μ襯 れ姶襦 
 螳 蠏
 觜 豕 =  覈 谿城 蟆
觜  蟆曙 螳覯 螻襴讀
 蟆曙螳覯
 朱 襷り覲襦  螳 譯殊伎覃 豐蠍  
螳 豕  覦レ朱 襷り覲襯 覲蟆渚 蟆 覦覲旧
襦  螻襴讀
  蠍一瑚鍵襯  覦レ朱 讌覃伎 覦覲旧朱 豕 
 蟇磯Μ 螳 螻 :  螳 襷り鍵 
 觜 覩碁 螳 : 蠍一瑚鍵 螻壱伎 覩襦
 襷り覲(W, b) 豐蠍 螳  W, b襯 願覃 觜
襯 豕 覲 螳 谿場碁!
觜  蟆曙 螳覯 螻襴讀
 蟆曙螳覯 れ 一
 蠏碁 蟲
 覦レ 2谿 螻糾 : W, b襦 蟲

 螳   讌 
  : 企 讌  れ姶
 一 : 覃伎  讌 
   れ姶襯 螳 讌
谿場 企!
 觜  蠍一瑚鍵襯 螻
 W b  ク覩碁 一
 蠍一瑚鍵 れ姶螳  螻褐 覦
レ 螳襴
螻襴讀 ろ
 data set 
 train 蟆曙螳覯  牛磯伎 
 train  覯 覦覲牛  W, b 谿剰鍵
#1 data set
#2 蟆曙 螳覯 覦覲牛  W, b 谿
蠍
 襴
  (learning rate)
 襦螳 螳 覦覲 襷 朱 蟆 企 蟆瑚襯 
 蟆覃   (豕螳 讌豎覯襴  )
 蟆覃 螳  (豕螳 谿城 一 覓 襷 覦覲旧 )
谿瑚襭
 襦 豌蟇語, 譟磯ゴ (覦伎 蟾), 觜覩碁眼
 瑚概讌 伎擦 螳 襭
 襦 覦煙  http://goo.gl/CNSdsB
 蟆曙螳覯 企語 - http://www.mathworks.com/matlabcentral/fx_files/27631/1/fff.png

More Related Content

What's hot (20)

Image Deep Learning る伎
Image Deep Learning る伎Image Deep Learning る伎
Image Deep Learning る伎
Youngjae Kim
螳覦襯 螻糾碁碁 tensor-flow
螳覦襯  螻糾碁碁 tensor-flow螳覦襯  螻糾碁碁 tensor-flow
螳覦襯 螻糾碁碁 tensor-flow
[ 襾語 ろ磯 - 語 襾語] 3. 覿襯
[ 襾語 ろ磯 - 語 襾語] 3. 覿襯[ 襾語 ろ磯 - 語 襾語] 3. 覿襯
[ 襾語 ろ磯 - 語 襾語] 3. 覿襯
Haesun Park
瑚概讌, 蠍郁 蠏碁Μ螻 ル
瑚概讌, 蠍郁 蠏碁Μ螻 ル瑚概讌, 蠍郁 蠏碁Μ螻 ル
瑚概讌, 蠍郁 蠏碁Μ螻 ル
Jinwon Lee
檎稼稼螳襴
檎稼稼螳襴檎稼稼螳襴
檎稼稼螳襴
譬 豕
A Beginner's guide to understanding Autoencoder
A Beginner's guide to understanding AutoencoderA Beginner's guide to understanding Autoencoder
A Beginner's guide to understanding Autoencoder
Lee Seungeun
讖讌襷ル2 - CNN RNN 覯
讖讌襷ル2 - CNN RNN 覯讖讌襷ル2 - CNN RNN 覯
讖讌襷ル2 - CNN RNN 覯
Modulabs
求== 求==梶求午メ求 8
求== 求==梶求午メ求 8求== 求==梶求午メ求 8
求== 求==梶求午メ求 8
Sunggon Song
Ai 蠏瑚願碓
Ai 蠏瑚願碓Ai 蠏瑚願碓
Ai 蠏瑚願碓
Tensorflow
TensorflowTensorflow
Tensorflow
chs71
Ch.5 Deep Learning
Ch.5 Deep LearningCh.5 Deep Learning
Ch.5 Deep Learning
PartPrime
1.Introduction to Python and TensorFlow
1.Introduction to Python and TensorFlow1.Introduction to Python and TensorFlow
1.Introduction to Python and TensorFlow
Haesun Park
瑚概 蟆暑 蟲 蟯 螳 る
瑚概 蟆暑 蟲 蟯 螳 る瑚概 蟆暑 蟲 蟯 螳 る
瑚概 蟆暑 蟲 蟯 螳 る
Woonghee Lee
4.representing data and engineering features
4.representing data and engineering features4.representing data and engineering features
4.representing data and engineering features
Haesun Park
R.T.Bach
R.T.BachR.T.Bach
R.T.Bach
豌 覦
襦磯 覦一磯 ル
襦磯 覦一磯 ル襦磯 覦一磯 ル
襦磯 覦一磯 ル
谿 譯
REALM
REALMREALM
REALM
Hoon Heo
伎 一危郁骸 覯2 - 一危 螳 れ 一危磯, 蠏碁Μ螻 襾語 覓 (2020 危)
伎 一危郁骸 覯2 - 一危 螳 れ 一危磯, 蠏碁Μ螻 襾語 覓 (2020 危)伎 一危郁骸 覯2 - 一危 螳 れ 一危磯, 蠏碁Μ螻 襾語 覓 (2020 危)
伎 一危郁骸 覯2 - 一危 螳 れ 一危磯, 蠏碁Μ螻 襾語 覓 (2020 危)
Tae Young Lee
2.supervised learning(epoch#2)-3
2.supervised learning(epoch#2)-32.supervised learning(epoch#2)-3
2.supervised learning(epoch#2)-3
Haesun Park
2.linear regression and logistic regression
2.linear regression and logistic regression2.linear regression and logistic regression
2.linear regression and logistic regression
Haesun Park
Image Deep Learning る伎
Image Deep Learning る伎Image Deep Learning る伎
Image Deep Learning る伎
Youngjae Kim
螳覦襯 螻糾碁碁 tensor-flow
螳覦襯  螻糾碁碁 tensor-flow螳覦襯  螻糾碁碁 tensor-flow
螳覦襯 螻糾碁碁 tensor-flow
[ 襾語 ろ磯 - 語 襾語] 3. 覿襯
[ 襾語 ろ磯 - 語 襾語] 3. 覿襯[ 襾語 ろ磯 - 語 襾語] 3. 覿襯
[ 襾語 ろ磯 - 語 襾語] 3. 覿襯
Haesun Park
瑚概讌, 蠍郁 蠏碁Μ螻 ル
瑚概讌, 蠍郁 蠏碁Μ螻 ル瑚概讌, 蠍郁 蠏碁Μ螻 ル
瑚概讌, 蠍郁 蠏碁Μ螻 ル
Jinwon Lee
檎稼稼螳襴
檎稼稼螳襴檎稼稼螳襴
檎稼稼螳襴
譬 豕
A Beginner's guide to understanding Autoencoder
A Beginner's guide to understanding AutoencoderA Beginner's guide to understanding Autoencoder
A Beginner's guide to understanding Autoencoder
Lee Seungeun
讖讌襷ル2 - CNN RNN 覯
讖讌襷ル2 - CNN RNN 覯讖讌襷ル2 - CNN RNN 覯
讖讌襷ル2 - CNN RNN 覯
Modulabs
求== 求==梶求午メ求 8
求== 求==梶求午メ求 8求== 求==梶求午メ求 8
求== 求==梶求午メ求 8
Sunggon Song
Ai 蠏瑚願碓
Ai 蠏瑚願碓Ai 蠏瑚願碓
Ai 蠏瑚願碓
Tensorflow
TensorflowTensorflow
Tensorflow
chs71
Ch.5 Deep Learning
Ch.5 Deep LearningCh.5 Deep Learning
Ch.5 Deep Learning
PartPrime
1.Introduction to Python and TensorFlow
1.Introduction to Python and TensorFlow1.Introduction to Python and TensorFlow
1.Introduction to Python and TensorFlow
Haesun Park
瑚概 蟆暑 蟲 蟯 螳 る
瑚概 蟆暑 蟲 蟯 螳 る瑚概 蟆暑 蟲 蟯 螳 る
瑚概 蟆暑 蟲 蟯 螳 る
Woonghee Lee
4.representing data and engineering features
4.representing data and engineering features4.representing data and engineering features
4.representing data and engineering features
Haesun Park
R.T.Bach
R.T.BachR.T.Bach
R.T.Bach
豌 覦
襦磯 覦一磯 ル
襦磯 覦一磯 ル襦磯 覦一磯 ル
襦磯 覦一磯 ル
谿 譯
伎 一危郁骸 覯2 - 一危 螳 れ 一危磯, 蠏碁Μ螻 襾語 覓 (2020 危)
伎 一危郁骸 覯2 - 一危 螳 れ 一危磯, 蠏碁Μ螻 襾語 覓 (2020 危)伎 一危郁骸 覯2 - 一危 螳 れ 一危磯, 蠏碁Μ螻 襾語 覓 (2020 危)
伎 一危郁骸 覯2 - 一危 螳 れ 一危磯, 蠏碁Μ螻 襾語 覓 (2020 危)
Tae Young Lee
2.supervised learning(epoch#2)-3
2.supervised learning(epoch#2)-32.supervised learning(epoch#2)-3
2.supervised learning(epoch#2)-3
Haesun Park
2.linear regression and logistic regression
2.linear regression and logistic regression2.linear regression and logistic regression
2.linear regression and logistic regression
Haesun Park

Similar to Howto_Tensorflow+Linear Regression (20)

Chapter 11 Practical Methodology
Chapter 11 Practical MethodologyChapter 11 Practical Methodology
Chapter 11 Practical Methodology
KyeongUkJang
ル Tensor flow(skt academy)
ル  Tensor flow(skt academy)ル  Tensor flow(skt academy)
ル Tensor flow(skt academy)
Tae Young Lee
2012 Ds 01
2012 Ds 012012 Ds 01
2012 Ds 01
Jungyerin
襦 蠍一 危危蠍
襦 蠍一 危危蠍 襦 蠍一 危危蠍
襦 蠍一 危危蠍
Yong Joon Moon
覲 襦蠏碁覦
覲 襦蠏碁覦覲 襦蠏碁覦
覲 襦蠏碁覦
[覦豺危] 手豌襴 螻豸
[覦豺危]  手豌襴 螻豸[覦豺危]  手豌襴 螻豸
[覦豺危] 手豌襴 螻豸
語伎 覦 危
語伎  覦 危語伎  覦 危
語伎 覦 危
譴 螻
Where to Apply Dropout in Recurrent Neural Networks for Handwriting Recognition?
Where to Apply Dropout in Recurrent Neural Networks for Handwriting Recognition?Where to Apply Dropout in Recurrent Neural Networks for Handwriting Recognition?
Where to Apply Dropout in Recurrent Neural Networks for Handwriting Recognition?
Lee Gyeong Hoon
AUTOML
AUTOMLAUTOML
AUTOML
Automl
AutomlAutoml
Automl
Exploring Deep Learning Acceleration Technology Embedded in LLMs
Exploring Deep Learning Acceleration Technology Embedded in LLMsExploring Deep Learning Acceleration Technology Embedded in LLMs
Exploring Deep Learning Acceleration Technology Embedded in LLMs
Tae Young Lee
, ろ碁 Api
, ろ碁 Api, ろ碁 Api
, ろ碁 Api
Chi Hwan Choi
伎一 Project7
伎一 Project7伎一 Project7
伎一 Project7
KoChungWook
Machine learning linearregression
Machine learning linearregressionMachine learning linearregression
Machine learning linearregression
HaYoungChoi17
[2B2]求 求メ釈梶 釈求戟梶 メ午梶= メ 求メ
[2B2]求 求メ釈梶 釈求戟梶 メ午梶= メ 求メ[2B2]求 求メ釈梶 釈求戟梶 メ午梶= メ 求メ
[2B2]求 求メ釈梶 釈求戟梶 メ午梶= メ 求メ
NAVER D2
螳覦螳 危危 覿
螳覦螳 危危 覿螳覦螳 危危 覿
螳覦螳 危危 覿
豌 覦
[225]NSML: 襾語 觜ろ蠍 & 覈 蠍
[225]NSML: 襾語  觜ろ蠍 & 覈  蠍[225]NSML: 襾語  觜ろ蠍 & 覈  蠍
[225]NSML: 襾語 觜ろ蠍 & 覈 蠍
NAVER D2
蟆讌 ChatGPT
蟆讌  ChatGPT蟆讌  ChatGPT
蟆讌 ChatGPT
Tae Young Lee
Lec 00, 01
Lec 00, 01Lec 00, 01
Lec 00, 01
Jinhwan Suk
TenforFlow Internals
TenforFlow InternalsTenforFlow Internals
TenforFlow Internals
Kiho Hong
Chapter 11 Practical Methodology
Chapter 11 Practical MethodologyChapter 11 Practical Methodology
Chapter 11 Practical Methodology
KyeongUkJang
ル Tensor flow(skt academy)
ル  Tensor flow(skt academy)ル  Tensor flow(skt academy)
ル Tensor flow(skt academy)
Tae Young Lee
2012 Ds 01
2012 Ds 012012 Ds 01
2012 Ds 01
Jungyerin
襦 蠍一 危危蠍
襦 蠍一 危危蠍 襦 蠍一 危危蠍
襦 蠍一 危危蠍
Yong Joon Moon
覲 襦蠏碁覦
覲 襦蠏碁覦覲 襦蠏碁覦
覲 襦蠏碁覦
[覦豺危] 手豌襴 螻豸
[覦豺危]  手豌襴 螻豸[覦豺危]  手豌襴 螻豸
[覦豺危] 手豌襴 螻豸
語伎 覦 危
語伎  覦 危語伎  覦 危
語伎 覦 危
譴 螻
Where to Apply Dropout in Recurrent Neural Networks for Handwriting Recognition?
Where to Apply Dropout in Recurrent Neural Networks for Handwriting Recognition?Where to Apply Dropout in Recurrent Neural Networks for Handwriting Recognition?
Where to Apply Dropout in Recurrent Neural Networks for Handwriting Recognition?
Lee Gyeong Hoon
AUTOML
AUTOMLAUTOML
AUTOML
Automl
AutomlAutoml
Automl
Exploring Deep Learning Acceleration Technology Embedded in LLMs
Exploring Deep Learning Acceleration Technology Embedded in LLMsExploring Deep Learning Acceleration Technology Embedded in LLMs
Exploring Deep Learning Acceleration Technology Embedded in LLMs
Tae Young Lee
伎一 Project7
伎一 Project7伎一 Project7
伎一 Project7
KoChungWook
Machine learning linearregression
Machine learning linearregressionMachine learning linearregression
Machine learning linearregression
HaYoungChoi17
[2B2]求 求メ釈梶 釈求戟梶 メ午梶= メ 求メ
[2B2]求 求メ釈梶 釈求戟梶 メ午梶= メ 求メ[2B2]求 求メ釈梶 釈求戟梶 メ午梶= メ 求メ
[2B2]求 求メ釈梶 釈求戟梶 メ午梶= メ 求メ
NAVER D2
螳覦螳 危危 覿
螳覦螳 危危 覿螳覦螳 危危 覿
螳覦螳 危危 覿
豌 覦
[225]NSML: 襾語 觜ろ蠍 & 覈 蠍
[225]NSML: 襾語  觜ろ蠍 & 覈  蠍[225]NSML: 襾語  觜ろ蠍 & 覈  蠍
[225]NSML: 襾語 觜ろ蠍 & 覈 蠍
NAVER D2
TenforFlow Internals
TenforFlow InternalsTenforFlow Internals
TenforFlow Internals
Kiho Hong

More from Hyo jeong Lee (10)

Project_Automatic Photo Classification Web Service
Project_Automatic Photo Classification Web ServiceProject_Automatic Photo Classification Web Service
Project_Automatic Photo Classification Web Service
Hyo jeong Lee
Progress_190118
Progress_190118Progress_190118
Progress_190118
Hyo jeong Lee
Progress_190130
Progress_190130Progress_190130
Progress_190130
Hyo jeong Lee
Progress_190213
Progress_190213Progress_190213
Progress_190213
Hyo jeong Lee
Progress_190412
Progress_190412Progress_190412
Progress_190412
Hyo jeong Lee
Progress_190315
Progress_190315Progress_190315
Progress_190315
Hyo jeong Lee
Paper_Scalable database logging for multicores
Paper_Scalable database logging for multicoresPaper_Scalable database logging for multicores
Paper_Scalable database logging for multicores
Hyo jeong Lee
Paper_An Efficient Garbage Collection in Java Virtual Machine via Swap I/O O...
Paper_An Efficient Garbage Collection in Java Virtual  Machine via Swap I/O O...Paper_An Efficient Garbage Collection in Java Virtual  Machine via Swap I/O O...
Paper_An Efficient Garbage Collection in Java Virtual Machine via Swap I/O O...
Hyo jeong Lee
Paper_Design of Swap-aware Java Virtual Machine Garbage Collector Policy
Paper_Design of Swap-aware Java Virtual Machine Garbage Collector PolicyPaper_Design of Swap-aware Java Virtual Machine Garbage Collector Policy
Paper_Design of Swap-aware Java Virtual Machine Garbage Collector Policy
Hyo jeong Lee
Howto_Tensorflow-slim
Howto_Tensorflow-slimHowto_Tensorflow-slim
Howto_Tensorflow-slim
Hyo jeong Lee
Project_Automatic Photo Classification Web Service
Project_Automatic Photo Classification Web ServiceProject_Automatic Photo Classification Web Service
Project_Automatic Photo Classification Web Service
Hyo jeong Lee
Paper_Scalable database logging for multicores
Paper_Scalable database logging for multicoresPaper_Scalable database logging for multicores
Paper_Scalable database logging for multicores
Hyo jeong Lee
Paper_An Efficient Garbage Collection in Java Virtual Machine via Swap I/O O...
Paper_An Efficient Garbage Collection in Java Virtual  Machine via Swap I/O O...Paper_An Efficient Garbage Collection in Java Virtual  Machine via Swap I/O O...
Paper_An Efficient Garbage Collection in Java Virtual Machine via Swap I/O O...
Hyo jeong Lee
Paper_Design of Swap-aware Java Virtual Machine Garbage Collector Policy
Paper_Design of Swap-aware Java Virtual Machine Garbage Collector PolicyPaper_Design of Swap-aware Java Virtual Machine Garbage Collector Policy
Paper_Design of Swap-aware Java Virtual Machine Garbage Collector Policy
Hyo jeong Lee
Howto_Tensorflow-slim
Howto_Tensorflow-slimHowto_Tensorflow-slim
Howto_Tensorflow-slim
Hyo jeong Lee

Howto_Tensorflow+Linear Regression

  • 2. Index ル螻 襦 襦 れ 蠏覿 覲螳 蟯螻 覈 觜 蟆曙 螳覯 螻襴讀 螻襴讀 ろ
  • 4. ル螻 襦 ル 蟆暑 螻襴讀 譯朱 殊碁 螻襴讀 CNN螻 RNN CNN(炎咳 蟆暑) 企語 伎 覓殊牡 語 RNN( 蟆暑) 覯螻 煙語 襦 豕蠏殊 ル 郁規 襴 殊企襴
  • 5. 襦 蠍磯蓋 蟲譟 一危 襦 蠏碁襯 豺 一一 . 蠏碁 碁 : 一 蠏碁 讌 : れ姶 一危 覦一 豺 一一 蠍壱碁 蠏碁 蟲譟磯ゼ 襷り 豌襴. レ CPU, GPU レ 覈 伎 螳 覈覦 , 64觜 襴れ 覈 伎 螳 覲企 覈 襷 覲企ゼ 覈磯螻 ろ 伎. 谿瑚 : https://goo.gl/6Cyr20
  • 6. 襦 vs 襦 觜 襦 ル 一危磯ゼ 螳讌螻 襾語 螻襴讀 襷 れ 覈語 貅譴 . 襦 觜 覈語 企殊伎 碁覿 れ伎る 一危磯ゼ 覦伎.
  • 8. #0 伎螻 伎宛 れ 伎 れ 伎 螳覦 蟆曙 伎宛 れ
  • 9. #1 docker れ 蟆 れ 襦 襴れ 覩襦 螳 襾語 譯朱 docker tool box襯 れ docker quick start terminal ろ docker-machine create vdocker d virtualbox 豕豐 1 ろ! docker-machine ls vdocker 螳 Running語 誤 Stopped企朱 docker-machine start vdocker 磯碁 譬襭
  • 10. #1 docker れ 蟆 れ
  • 11. #2 cmd docker 企語 れ企 cmd ろ 覈 (伎郁鍵 譯殊.) FOR /f "tokens=*" %i IN ('docker-machine env --shell cmd default') DO %i docker 企語 れ企 (豕豐 1 ろ!) docker run v /c/Users/ML:/home/ML it python bash 伎宛 襦 蟆暑襯 襦 譯手鍵 焔 企語覈 誤蠍 : docker ps a : docker ps rm (image name) る docker stop (image name) 螻 覈
  • 12. #3 焔 貉企 覦 docker restart (image name) docker attach (image name) + Enter (所 覯蠏!) 覃 docker start (image name) attach ろ
  • 13. #4 襦 れ https://www.tensorflow.org/versions/r0.11/get_started/os_set up.html 貉危 蟆曙 襷 覈轟 襦 れ 襭覃 襦 启襦 企 襦碁ゼ ろ企蓋! cd (path) python (file name) 朱 覿覿蟾讌 ロ螻 Tabるゼ 襯企 企 朱
  • 14. #5 れ 襭 れ ロ 螻殊
  • 15. #6 螳 ろ : 螻襴讀 蠍一 > 語 > 一 ろ! placeholder 襦蠏碁 ろ 譴 螳 覲蟆渚 覯襴 覲 tensor 蠍 螳 れ姶 一危 覦 mul() 譟一 襦螳 螻牛 襷 一 譴 Session() 語 燕朱 襦蠏碁 襦 殊企襴 語
  • 16. Note that 襦 一郁骸 一危一 覈 覲企 蠏碁 蟲譟 ル 碁 : 一 一危 リ骸 豢レ 豺襯 蟇磯 ル 覲襯 所 讌 : 螳螻 豢 螳朱 郁屋 碁 螳 蟯螻襯 + 襦 蠍磯蓋 襭蟲譟一 企!
  • 17. 蠏 覿 覲 螳 蟯螻 覈 觜 蟆曙 螳覯 螻襴讀 螻襴讀 ろ
  • 18. 蠏 覿(linear regression)企? 覲 伎 蟯螻襯 覿 糾 覦 覯 襴暑 xi, b 譬覲 y螳 蟯螻襯 覈碁 蠏 れ蠏 蠏 : 覲 螳 蟯螻 蟆曙 れ蠏 : 螳 覲襯 る 蟆曙
  • 19. 覲 螳 蟯螻 覈 : y = W*x+b 伎姶 譬螻 一危磯ゼ 燕 伎 襦蠏碁 一危一 襷 豕 讌 谿城 覓語
  • 20. 觜 (cost function) 蟆曙 螳覯 螻襴讀(gradient descent) 覓語 : 一危磯覿 豢レ 豢 螻襴讀 り鍵 W b 螳 襷り覲襦 覈 螳 一危 x_data襯 伎 豢 一危 y_data襯 襷 豕 襷り覲 W b襯 谿剰鍵
  • 21. 觜 蟆曙 螳覯 螻襴讀 觜(=れ姶) 覦覲旧 殊企 襷 螳 覿襯 誤蠍 朱 譬 () 讌語 豸′ W b襯 襷り覲襦 覦 讌 朱 一危一 襷讌襯 蠍一朱 れ姶 螳 襴 蠍一 觜襦 蠏螻煙れ姶(mean square error) れ 螳螻 螻襴讀 覦覲給 豢 螳 伎 蟇磯Μ襯 れ姶襦 螳 蠏 觜 豕 = 覈 谿城 蟆
  • 22. 觜 蟆曙 螳覯 螻襴讀 蟆曙螳覯 朱 襷り覲襦 螳 譯殊伎覃 豐蠍 螳 豕 覦レ朱 襷り覲襯 覲蟆渚 蟆 覦覲旧 襦 螻襴讀 蠍一瑚鍵襯 覦レ朱 讌覃伎 覦覲旧朱 豕 蟇磯Μ 螳 螻 : 螳 襷り鍵 觜 覩碁 螳 : 蠍一瑚鍵 螻壱伎 覩襦 襷り覲(W, b) 豐蠍 螳 W, b襯 願覃 觜 襯 豕 覲 螳 谿場碁!
  • 23. 觜 蟆曙 螳覯 螻襴讀 蟆曙螳覯 れ 一 蠏碁 蟲 覦レ 2谿 螻糾 : W, b襦 蟲 螳 讌 : 企 讌 れ姶 一 : 覃伎 讌 れ姶襯 螳 讌 谿場 企! 觜 蠍一瑚鍵襯 螻 W b ク覩碁 一 蠍一瑚鍵 れ姶螳 螻褐 覦 レ 螳襴
  • 24. 螻襴讀 ろ data set train 蟆曙螳覯 牛磯伎 train 覯 覦覲牛 W, b 谿剰鍵
  • 26. #2 蟆曙 螳覯 覦覲牛 W, b 谿 蠍
  • 27. (learning rate) 襦螳 螳 覦覲 襷 朱 蟆 企 蟆瑚襯 蟆覃 (豕螳 讌豎覯襴 ) 蟆覃 螳 (豕螳 谿城 一 覓 襷 覦覲旧 )
  • 28. 谿瑚襭 襦 豌蟇語, 譟磯ゴ (覦伎 蟾), 觜覩碁眼 瑚概讌 伎擦 螳 襭 襦 覦煙 http://goo.gl/CNSdsB 蟆曙螳覯 企語 - http://www.mathworks.com/matlabcentral/fx_files/27631/1/fff.png

Editor's Notes

  • #14: apt-get install python-pip python-dev export TF_BINARY_URL=https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow-0.11.0-cp35-cp35m-linux_x86_64.whl pip install --upgrade $TF_BINARY_URL