hard

http://blog.livedoor.jp/wisteriear/archives/1062055501.html

https://www.pc-koubou.jp/casestudy/gpgpu.php

语音识别

https://www.zhihu.com/question/20398418

https://github.com/itzikgili/speech-to-text-wavenet

iris segmentation

http://islab.hh.se/mediawiki/Iris_Segmentation_Code_Based_on_the_GST

dicomdata

https://www.kaggle.com/c/data-science-bowl-2017/data

tensorflow with batch normalization

http://r2rt.com/implementing-batch-normalization-in-tensorflow.html

フレームワーク比較

http://geek.csdn.net/news/detail/138968

softplus and ReLU

http://www.cnblogs.com/neopenx/p/4453161.html

雅可比行列式和海森矩阵

http://jacoxu.com/jacobian%E7%9F%A9%E9%98%B5%E5%92%8Chessian%E7%9F%A9%E9%98%B5/

matplotlib

import numpy as np
from pandas import *
from pylab import *
import matplotlib.pyplot as plt
from matplotlib import font_manager
from numpy.random import randn

def sigmoid(x):
    return 1.0/(1.0 + np.exp(-x))

def sigmoid_prime(x):
    return sigmoid(x)*(1-sigmoid(x))

# 日本語を使うため必要
fontprop = matplotlib.font_manager.FontProperties(fname="/usr/share/fonts/truetype/fonts-japanese-gothic.ttf")

# 新規のウィンドウを描画
fig = plt.figure()

# サブプロットを追加
#ax1 = fig.add_subplot(2,2,1)
ax2 = fig.add_subplot(1,2,1)
ax3 = fig.add_subplot(1,2,2)

x = np.arange(-3, 3, 0.1)
y = np.sin(x)
#ax1.plot(x, y)

y1 = sigmoid(x)
ax2.plot(x, y1)

y2 = sigmoid_prime(x)
ax3.plot(x, y2)

plt.show()
plt.savefig("image.png")

adam実装

https://github.com/pfnet/chainer/tree/master/chainer/optimizers

http://blog.csdn.net/csyanbin/article/details/53460627

https://github.com/fchollet/keras/blob/master/keras/optimizers.py

http://www.2cto.com/kf/201611/562392.html

http://blog.csdn.net/heyongluoyao8/article/details/52478715

cnn free book state of art

http://lamda.nju.edu.cn/zhangh/Convolutional%20Neuron%20Networks%20and%20its%20Applications.pdf

video lecture

Nando de Freitas

data set for cancer

https://archive.ics.uci.edu/ml/datasets/Breast+Cancer+Wisconsin+%28Diagnostic%29

matplotlib使用方法

import numpy as np
from pandas import *
from pylab import *
import matplotlib.pyplot as plt
from matplotlib import font_manager
from numpy.random import randn

# 日本語を使うため必要
fontprop = matplotlib.font_manager.FontProperties(fname="/usr/share/fonts/truetype/fonts-japanese-gothic.ttf")

# 新規のウィンドウを描画
fig = plt.figure()

# サブプロットを追加
ax1 = fig.add_subplot(2,2,1)
ax2 = fig.add_subplot(2,2,2)
ax3 = fig.add_subplot(2,2,3)

plt.show()
plt.savefig("image.png")

SGD算法比较

https://blog.slinuxer.com/2016/09/sgd-comparison

http://geek.csdn.net/news/detail/109739

dl paper reading roadmap

https://github.com/songrotek/Deep-Learning-Papers-Reading-Roadmap

matplotlib

http://qiita.com/ynakayama/items/8d3b1f7356da5bcbe9bc

rcnn

https://www.zhihu.com/question/35887527

http://insidehpc.com/2017/01/intel-fpgas-break-record-deep-learning-facial-recognition/

http://pjreddie.com/darknet/yolo/

GANｓ

http://musyoku.github.io/2016/10/28/Deep-Directed-Generative-Models-with-Energy-Based-Probability-Estimation/

http://sotetsuk.hatenablog.com/entry/2015/12/16/185102

https://arxiv.org/pdf/1412.6572v3.pdf

http://geek.csdn.net/news/detail/134553

simGANソース
https://github.com/wayaai/SimGAN

资料整理
http://blog.csdn.net/solomon1558/article/details/52338052

论文集
https://github.com/zhangqianhui/AdversarialNetsPapers

https://zhuanlan.zhihu.com/p/25071913?utm_medium=social&utm_source=wechat_timeline&from=singlemessage&isappinstalled=0

http://mp.weixin.qq.com/s/trvMOTXNs7L6fSmTkZXwsA