<?xml version="1.0" encoding="utf-8" ?><rss version="2.0"><channel><title><![CDATA[jiandanjinxin的专栏]]></title><description><![CDATA[http://www.jianshu.com/users/8ad7903302b3/timeline 有时候抄写学习笔记，请勿见怪]]></description><link>https://blog.youkuaiyun.com/jiandanjinxin</link><language>zh-cn</language><generator>https://blog.youkuaiyun.com/</generator><copyright><![CDATA[Copyright &copy; jiandanjinxin]]></copyright><item><title><![CDATA[failed to initialize nvml driver/library version mismatch ubuntu]]></title><link>https://blog.youkuaiyun.com/jiandanjinxin/article/details/80688900</link><guid>https://blog.youkuaiyun.com/jiandanjinxin/article/details/80688900</guid><author>jiandanjinxin</author><pubDate>Thu, 14 Jun 2018 10:34:00 +0800</pubDate><description><![CDATA[英伟达驱动版本是384.130显示的NVRM version: NVIDIA UNIX x86_64 Kernel Module是：384.130。 若是旧的版本就会出现如下问题。这个问题出现的原因是kernel mod 的 Nvidia driver 的版本没有更新，一般情况下，重启机器就能够解决，如果因为某些原因不能够重启的话，也有办法reload kernel mod。方法如下：就两步unl...]]></description><category></category></item><item><title><![CDATA[Cheat_Sheet ---Keras、Matlab、Matplotlib、Numpy、Pandas、Scikit-Learn、SciPy]]></title><link>https://blog.youkuaiyun.com/jiandanjinxin/article/details/80002351</link><guid>https://blog.youkuaiyun.com/jiandanjinxin/article/details/80002351</guid><author>jiandanjinxin</author><pubDate>Thu, 19 Apr 2018 11:30:12 +0800</pubDate><description><![CDATA[Cheat_Sheet ---KerasCheat_Sheet ---MatlabCheat_Sheet ---MatplotlibCheat_Sheet ---NumpyCheat_Sheet ---PandasCheat_Sheet ---Scikit-LearnCheat_Sheet ---SciPy参考文献http://ddl.escience.cn/f/IDkq#path=%2F8215...]]></description><category></category></item><item><title><![CDATA[How to plot scikit learn classification report?]]></title><link>https://blog.youkuaiyun.com/jiandanjinxin/article/details/79731177</link><guid>https://blog.youkuaiyun.com/jiandanjinxin/article/details/79731177</guid><author>jiandanjinxin</author><pubDate>Wed, 28 Mar 2018 17:38:04 +0800</pubDate><description><![CDATA[````import matplotlib.pyplot as pltimport numpy as npdef show_values(pc, fmt=&quot;%.2f&quot;, **kw):    '''    Heatmap with text in each cell with matplotlib's pyplot    Source: https://stackoverflow.com/a/250...]]></description><category></category></item><item><title><![CDATA[Google Colab 免费GPU服务器使用教程]]></title><link>https://blog.youkuaiyun.com/jiandanjinxin/article/details/79680976</link><guid>https://blog.youkuaiyun.com/jiandanjinxin/article/details/79680976</guid><author>jiandanjinxin</author><pubDate>Sat, 24 Mar 2018 20:58:15 +0800</pubDate><description><![CDATA[Google Colab免费使用GPU服务器教程Google Colab Free GPU TutorialColaboratory相关查询命令```!/opt/bin/nvidia-smi``````!pip install!apt-get install!pip install -q kerasimport keras!apt-get install libnvrtc8.0!pip insta...]]></description><category></category></item><item><title><![CDATA[Python xrange与range的区别]]></title><link>https://blog.youkuaiyun.com/jiandanjinxin/article/details/78849794</link><guid>https://blog.youkuaiyun.com/jiandanjinxin/article/details/78849794</guid><author>jiandanjinxin</author><pubDate>Wed, 20 Dec 2017 09:57:48 +0800</pubDate><description><![CDATA[xrange 与 rangexrange 用法与 range 完全相同，所不同的是生成的不是一个list对象，而是一个生成器。要生成很大的数字序列的时候，用xrange会比range性能优很多，因为不需要一上来就开辟一块很大的内存空间。 
xrange 和 range 这两个基本上都是在循环的时候用。for i in range(0, 100):
    print ifor i in xrange]]></description><category></category></item><item><title><![CDATA[画图网页http://weavesilk.com/]]></title><link>https://blog.youkuaiyun.com/jiandanjinxin/article/details/78846388</link><guid>https://blog.youkuaiyun.com/jiandanjinxin/article/details/78846388</guid><author>jiandanjinxin</author><pubDate>Tue, 19 Dec 2017 19:44:15 +0800</pubDate><description><![CDATA[画图网页http://weavesilk.com/]]></description><category></category></item><item><title><![CDATA[word_cloud]]></title><link>https://blog.youkuaiyun.com/jiandanjinxin/article/details/78801521</link><guid>https://blog.youkuaiyun.com/jiandanjinxin/article/details/78801521</guid><author>jiandanjinxin</author><pubDate>Thu, 14 Dec 2017 13:18:36 +0800</pubDate><description><![CDATA[word_cloudword_cloudwordcloud_server]]></description><category></category></item><item><title><![CDATA[李宏毅机器学习课程－Structured Learning]]></title><link>https://blog.youkuaiyun.com/jiandanjinxin/article/details/78791699</link><guid>https://blog.youkuaiyun.com/jiandanjinxin/article/details/78791699</guid><author>jiandanjinxin</author><pubDate>Wed, 13 Dec 2017 15:39:24 +0800</pubDate><description><![CDATA[Simple structured learning framework for python 
pystruct-github 
Slides for explaining structured prediction and PyStruct -github



一、Structured Learning-Unifed Framework







之前的input and output]]></description><category></category></item><item><title><![CDATA[李宏毅机器学习课程－Transfer Learning]]></title><link>https://blog.youkuaiyun.com/jiandanjinxin/article/details/78787878</link><guid>https://blog.youkuaiyun.com/jiandanjinxin/article/details/78787878</guid><author>jiandanjinxin</author><pubDate>Wed, 13 Dec 2017 09:41:30 +0800</pubDate><description><![CDATA[深度学习 -> 强化学习 ->迁移学习（杨强教授报告)

李宏毅机器学习课程－Transfer Learning

迁移学习－吴恩达 
freeze





待处理的

理解深层神经网络中的迁移学习及TensorFlow实现 



Transfer Learning模式



Similar domain, different tasks



Different domains, same]]></description><category></category></item><item><title><![CDATA[12bit灰度图像映射到8bit显示及python 实现]]></title><link>https://blog.youkuaiyun.com/jiandanjinxin/article/details/78365183</link><guid>https://blog.youkuaiyun.com/jiandanjinxin/article/details/78365183</guid><author>jiandanjinxin</author><pubDate>Fri, 27 Oct 2017 15:11:15 +0800</pubDate><description><![CDATA[图像显示和打印面临的一个问题是：图像的亮度和对比度能否充分突出关键部分。这里所指的“关键部分”在   CT   里的例子有软组织、骨头、脑组织、肺、腹部等等。  技术问题1、显示器往往只有   8-bit，   而数据有   12-   至   16-bits。 
2、如果将数据的   min   和   max   间   (dynamic   range)   的之间转换到   8-bit]]></description><category></category></item><item><title><![CDATA[应用深度学习(台大陈蕴侬&李宏毅) Part1]]></title><link>https://blog.youkuaiyun.com/jiandanjinxin/article/details/78359025</link><guid>https://blog.youkuaiyun.com/jiandanjinxin/article/details/78359025</guid><author>jiandanjinxin</author><pubDate>Thu, 26 Oct 2017 21:35:20 +0800</pubDate><description><![CDATA[History of Deep LearningBig Data  & GPU端到端Universality TheoremCore Factors for Applied Deep Learning参考文献http://v.qq.com/vplus/578e2d6f5e1fadc1/foldervideos/8n1000201qzzkx5Deep Learning 
◦Goodfellow, Be]]></description><category></category></item><item><title><![CDATA[关于Matlab编程的思考（待续）]]></title><link>https://blog.youkuaiyun.com/jiandanjinxin/article/details/78353027</link><guid>https://blog.youkuaiyun.com/jiandanjinxin/article/details/78353027</guid><author>jiandanjinxin</author><pubDate>Thu, 26 Oct 2017 13:57:08 +0800</pubDate><description><![CDATA[Matlab编程的规范化思考

1.并行化

2.释放内存

3.需要调参的变量太多，可考虑将变量都放到一个结构体里面。

4.find(y),就是要找到y中那些非零项的指引

5.代码运行出现问题的时候，在命令行输入why就可以得到答案

6.输入bench可以给电脑跑分。

7.home 将光标移至命令窗口的左上角

8.查看matlab命令历史记录可找history.m,



 prefd...]]></description><category></category></item><item><title><![CDATA[python中的深拷贝与浅拷贝]]></title><link>https://blog.youkuaiyun.com/jiandanjinxin/article/details/78214133</link><guid>https://blog.youkuaiyun.com/jiandanjinxin/article/details/78214133</guid><author>jiandanjinxin</author><pubDate>Thu, 12 Oct 2017 14:00:46 +0800</pubDate><description><![CDATA[浅拷贝的时候，修改原来的对象，深拷贝的对象不会发生改变。对象的赋值对象的赋值实际上是对象之间的引用：当创建一个对象，然后将这个对象赋值给另外一个变量的时候，python并没有拷贝这个对象，而只是拷贝了这个对象的引用。aList = ["kel","abc",123]
print(aList, id(aList))
bList = aListbList.append("add")print(aList]]></description><category></category></item><item><title><![CDATA[Python程序设计—车万翔]]></title><link>https://blog.youkuaiyun.com/jiandanjinxin/article/details/78196914</link><guid>https://blog.youkuaiyun.com/jiandanjinxin/article/details/78196914</guid><author>jiandanjinxin</author><pubDate>Tue, 10 Oct 2017 20:15:53 +0800</pubDate><description><![CDATA[程序设计入门—Python对象和类型五种基本对象类型字符串 （string），简记为 str 
使用 ’ ’ 或 ” ” 括起来的一系列字符 
 整数（integer），简记为 int 
 十进制：21，八进制：025，十六进制：0x15 
 浮点数（float） 
 1.48，21.0，21.，.21，2.1E2 
布尔数（boolean），简记为 bool 
True，False 
 复数（co]]></description><category></category></item><item><title><![CDATA[Tensor数据相关的运算、函数讲解及与numpy区别]]></title><link>https://blog.youkuaiyun.com/jiandanjinxin/article/details/78084390</link><guid>https://blog.youkuaiyun.com/jiandanjinxin/article/details/78084390</guid><author>jiandanjinxin</author><pubDate>Mon, 25 Sep 2017 13:57:50 +0800</pubDate><description><![CDATA[Tensortensorflow 中使用它来表示数据。可以看做多维数组或者list。 
标量是张量，向量是张量，矩阵是张量，矩阵的矩阵是张量。常用几种定义方法1. variable变量，一般是可以被更更新或更改的数值，即在流图运行过程中可以被不断动态调整的值。我们训练一个模型的时候，会用到Tensorflow中的变量(Variables)，我们需要它来保持和更新参数值，和张量一样，变量也保存在内存缓]]></description><category></category></item><item><title><![CDATA[李宏毅机器学习课程12~~~半监督学习]]></title><link>https://blog.youkuaiyun.com/jiandanjinxin/article/details/77961247</link><guid>https://blog.youkuaiyun.com/jiandanjinxin/article/details/77961247</guid><author>jiandanjinxin</author><pubDate>Wed, 13 Sep 2017 07:59:58 +0800</pubDate><description><![CDATA[Semi-supervised LearningThe distribution of the unlabeled data tell us something. Usually with some assumptions.Supervised Generative Model对比学习见 李宏毅机器学习课程４~~~分类：概率生成模型EM算法思路来最大化似然函数。Self-training 
Self]]></description><category></category></item><item><title><![CDATA[李宏毅机器学习课程11~~~为何要"深"？]]></title><link>https://blog.youkuaiyun.com/jiandanjinxin/article/details/77940801</link><guid>https://blog.youkuaiyun.com/jiandanjinxin/article/details/77940801</guid><author>jiandanjinxin</author><pubDate>Tue, 12 Sep 2017 08:46:30 +0800</pubDate><description><![CDATA[为何要“深”？ 
pluskid的博客 Deep Learning and Shallow LearningBengio Y. Learning deep architectures for AI. Foundations and trends® in Machine Learning, 2009Deeper is Better?模型有更多的参数会有更好的结果，这是毋庸置疑的。深瘦的模型会比浅胖的模]]></description><category></category></item><item><title><![CDATA[ubuntu16.04 制作gif]]></title><link>https://blog.youkuaiyun.com/jiandanjinxin/article/details/77890475</link><guid>https://blog.youkuaiyun.com/jiandanjinxin/article/details/77890475</guid><author>jiandanjinxin</author><pubDate>Fri, 08 Sep 2017 08:28:04 +0800</pubDate><description><![CDATA[byzanz安装sudo apt-get install byzanz    byzanz-record   #录像
    byzanz-playback   #回放下载完成后打开命令行输入byzanz-record –help其中我们重点关注几个参数
* -d 动画录制的时间,默认录制10秒
* -e 动画开始延迟
* -x 录制区域的起始X坐标
* -y 录制区域的起始Y坐标
* -w 录制区]]></description><category></category></item><item><title><![CDATA[github访问太慢解决方案]]></title><link>https://blog.youkuaiyun.com/jiandanjinxin/article/details/77774050</link><guid>https://blog.youkuaiyun.com/jiandanjinxin/article/details/77774050</guid><author>jiandanjinxin</author><pubDate>Fri, 01 Sep 2017 15:51:19 +0800</pubDate><description><![CDATA[github 域名列表先建立一个域名列表haha.txt，下面列表中的gist.github.com是代码片功能github.com
assets-cdn.github.com
avatars0.githubusercontent.com
avatars1.githubusercontent.com
documentcloud.github.com
gist.github.com
help.gith]]></description><category></category></item><item><title><![CDATA[Python中的除法保留两位小数]]></title><link>https://blog.youkuaiyun.com/jiandanjinxin/article/details/77752297</link><guid>https://blog.youkuaiyun.com/jiandanjinxin/article/details/77752297</guid><author>jiandanjinxin</author><pubDate>Thu, 31 Aug 2017 20:30:43 +0800</pubDate><description><![CDATA[在C/C++语言对于整形数执行除法会进行地板除（舍去小数部分）。例如 int a=15/10; a的结果为1。同样的在Java中也是如此，所以两个int型的数据相除需要返回一个浮点型数据的时候就需要强制类型转换，例如 float a = (float)b/c ,其中b、c都是int型数据。Python中分为3种除法：传统除法、精确除法、地板除。传统除法如果是整数除法则执行地板除，如果是浮点数除法则执]]></description><category></category></item></channel></rss>