Google的Tensorflow tutorial中,有提到這個連結
從最基本的觀念開始說起,內容含括SGD, back propagation, 不同activation function及cost function的探討,最後以CNN做為篇章的結尾
http://neuralnetworksanddeeplearning.com/index.html
圖示很清楚的一個ML網站
什麼是SVM?
http://cs231n.github.io/linear-classify/
動畫說明Linear Classification
http://vision.stanford.edu/teaching/cs231n-demos/linear-classify/
史丹佛一門介紹CNN(Convolutional Neural Network)的課程,有清楚的動畫講解
對於一般最容易誤解的depth colum(fibre)或parameter sharing,說明的很清楚
http://cs231n.github.io/convolutional-networks/#architectures
TensorFlow上如何做debug
https://wookayin.github.io/tensorflow-talk-debugging
如何利用Machine Learning來作畫
https://www.slideshare.net/ckmarkohchang/a-neural-algorithm-of-artistic-style
如何在CCTV上面,利用temporal reference及scene-specific reference來加速machine learning
辨別obect的速度
NoScope: 1000x Faster Deep Learning Queries over Video