5월, 2018의 게시물 표시

[Deep learning] Gradient Descent Optimizer

이미지
https://icim.nims.re.kr/post/easyMath/70 알기 쉬운 산업수학 알기 쉬운 산업수학 Gradient Descent Optimizer 2017년 12월 13일(수) | 김민중 URL  (1) 주어진 목적함수  f f 의 최솟값(minimum)을 찾아가는 알고리즘으로 다음과 같은 방식으로 최솟값을 찾아간다. 초기값  x 0 x 0 을 임의로 설정, 적당한 Learning rate  α α  설정 n ≥ 0 n ≥ 0 인 정수에 대해서  x n + 1 x n + 1 은 다음과 같이 정의한다. x n + 1 : = x n − α ⋅ ∇ f ( x n ) x n + 1 := x n − α ⋅ ∇ f ( x n ) 주의사항 함수 f f 의 모양이 convex가 아닌 경우 global minimum이 아닌 local minimum으로  x n x n 이 수렴할 가능성이 있다. Learning rate  α α  값이 큰 경우 최솟값으로  x n x n 이 수렴하는 것이 아니라 발산할 수 있다. Learning rate  α α  값이 작은 경우 수렴하는 속도가 지나치게 느릴 수 있다. 방정식  2 ⋅ x = 10 2 ⋅ x = 10  의 근을 Gradient Descent를 이용해서 찾아보자. 목적함수  f ( x ) : = ( 10 − 2 x ) 2 f ( x ) := ( 10 − 2 x ) 2 으로 설정하겠다( f f 의 최솟값인 0이 되게 하는  x x 값이 우리가 원하는 방정식의 근이다). 초기값  x 0 = 0 x 0 = 0 , Learning rate  α = 0.05 α = 0.05 으로 설정 f ′ ( x ) = 4 ( 2 x − 10 ) f ′ ( x ) = 4 ( 2 x − 10 ) 이므로  x 1 x 1 은 다음과 같이 구할 수 있다. x 1 = x 0 − α ⋅ f ′ ( x 0 ) = 0 − 0.05 ⋅ ( − 40 ) = 2 x

[Tensorflow][Mobile][Windows] TensorFlow for Poets 2: TFMobile

1. Install TensorFlow https://sunggukim.blogspot.kr/2018/05/tensorflow-install-tensorflow-with.html 2. download source for TFMobile for Poets2 git clone https://github.com/googlecodelabs/tensorflow-for-poets-2 cd tensorflow-for-poets-2 git checkout end_of_first_codelab dir tf_files/ 3. test model python -m scripts.label_image \ --graph=tf_files/retrained_graph.pb \ --image=tf_files/flower_photos/daisy/3475870145_685a19116d.jpg 4. Optimize model python -m tensorflow.python.tools.optimize_for_inference \ --input=tf_files/retrained_graph.pb \ --output=tf_files/optimized_graph.pb \ --input_names="input" \ --output_names="final_result" 5. Quantize weights python -m scripts.quantize_graph \ --input=tf_files/optimized_graph.pb \ --output=tf_files/rounded_graph.pb \ --output_node_names=final_result \ --mode=weights_rounded 6. Check compress rate gzip -c tf_files/rounded_graph.pb > tf_f

[Tensorflow] install tensorflow with anaconda 3.5 [ python 3.5]

이미지
1. download anaconda3.5 https://repo.continuum.io/archive/index.html ** my case : for windows x64 : https://repo.continuum.io/archive/Anaconda3-4.2.0-Windows-x86_64.exe 2. install anaconda ** my case : C:\Users\[user-account]\Anaconda3 3. Run anaconda prompt 4. Create anaconda virtual environment for tensorflow conda create -n tensorflow python=3.5 5. Activate tensorflow virtualenvironment activate tensorflow 6. Upgrade pip version python -m pip install --upgrade pip 7. Install tensorflow pip install tensorflow 8. Test and confirm python import tensorflow as tf hello = tf.constant("Hello tensorflow!") sess = tf.Session() print(sess.run(hello)) ** output :  b'Hello tensorflow!'

[AWS][Ubuntu AMI] install virtualenv virtualenvwrapper

0. Set root password $ su passwd and enter password two times. 1. Install pip, virtualenv , virtualenvwrapper for separate python environments $ wget https://bootstrap.pypa.io/get-pip.py $ sudo python get-pip.py $ sudo pip install virtualenv virtualenvwrapper $ sudo rm -rf get-pip.py ~/.cache/pip 2. Update .bashrc file [ nano ~/.bashrc ] *** we are going to use python3 because AWS Ubuntu AMI has no default python command # virtualenv and virtualenvwrapper export WORKON_HOME=$HOME/.virtualenvs export VIRTUALENVWRAPPER_PYTHON=/usr/bin/python3 source /usr/local/bin/virtualenvwrapper.sh 3. Make workplace with virtual environment $ source ~/.bashrc $ mkvirtualenv myVirtualEnv ( myVirtualEnv is name , it means you can name it , after this command run prompt of terminal will change. ) <Reference> https://medium.com/@gitudaniel/installing-virtualenvwrapper-for-python3-ad3dfea7c717

[Blogger][Syntax Highlight] How to syntax highlight in blogger

<HowTo> Edit template Insert bleow lines between <head> and </head> <!-- INCLUDE SyntaxHighlighter Components (css and javascript) --> <link href='http://alexgorbatchev.com/pub/sh/current/styles/shCore.css' rel='stylesheet' type='text/css'/> <link href='http://alexgorbatchev.com/pub/sh/current/styles/shThemeDefault.css' rel='stylesheet' type='text/css'/> <script src='http://alexgorbatchev.com/pub/sh/current/scripts/shCore.js' type='text/javascript'/> <!-- END INCLUDE SyntaxHighlighter Components --> <!-- INCLUDE YOUR BRUSHES WHICH YOU WANT TO USE HERE --> <script src='http://alexgorbatchev.com/pub/sh/current/scripts/shBrushCpp.js' type='text/javascript'/> <script src='http://alexgorbatchev.com/pub/sh/current/scripts/shBrushCSharp.js' type='text/javascript'/> <script src='http://alexgorbatchev.com/pub/sh/curre

[Game] Cute Monsters Brick Breaker, Classic Ball Game

이미지
https://appadvice.com/app/cute-monsters-brick-breaker-classic-ball-game/1230844281