<?xml version="1.0" encoding="utf-8" ?><rss version="2.0"><channel><title><![CDATA[KangRoger的专栏]]></title><description><![CDATA[]]></description><link>https://blog.youkuaiyun.com/KangRoger</link><language>zh-cn</language><generator>https://blog.youkuaiyun.com/</generator><copyright><![CDATA[Copyright &copy; KangRoger]]></copyright><item><title><![CDATA[容器、迭代器和算法]]></title><link>https://blog.youkuaiyun.com/KangRoger/article/details/83515068</link><guid>https://blog.youkuaiyun.com/KangRoger/article/details/83515068</guid><author>KangRoger</author><pubDate>Mon, 29 Oct 2018 23:15:49 +0800</pubDate><description><![CDATA[文章目录容器、迭代器和算法容器的实现Java的实现方法C++的实现方法容器与迭代器链表容器与迭代器集合与迭代器迭代器与算法求容器中元素纸盒微型算法库容器和迭代器的分类容器的陷阱
这是《深入实践C++模板编程》第五章“容器、迭代器和算法”的读书笔记。
容器、迭代器和算法
通过C++模板可以将类型以及其他编译期常数作为参数抽离出来，使代码拜托对类型依赖，从而设计容纳不同类型的容器成为可能。

容器是指...]]></description><category></category></item><item><title><![CDATA[模板特例]]></title><link>https://blog.youkuaiyun.com/KangRoger/article/details/83020606</link><guid>https://blog.youkuaiyun.com/KangRoger/article/details/83020606</guid><author>KangRoger</author><pubDate>Thu, 11 Oct 2018 23:32:45 +0800</pubDate><description><![CDATA[文章目录模板特例vector\模板特例
通过函数模板和类模板，可以为不同类型数据编写统一函数和类。但是现实情况比想象复杂，单一模板很难兼容各种情况。C++还提供了模板特例（template partial specialization）：对于某个已有模板，可以为某个或某组模板参数类型另外一种一种变体。
vector&amp;amp;amp;lt;bool&amp;amp;amp;gt;
模板特例的一个例子是STL中的vector&amp;amp;amp;lt;T&amp;amp;amp;gt...]]></description><category></category></item><item><title><![CDATA[模板参数类型]]></title><link>https://blog.youkuaiyun.com/KangRoger/article/details/82833001</link><guid>https://blog.youkuaiyun.com/KangRoger/article/details/82833001</guid><author>KangRoger</author><pubDate>Mon, 24 Sep 2018 22:26:36 +0800</pubDate><description><![CDATA[文章目录模板参数类型整数模板参数函数指针模板参数指针及引用模板参数成员函数指针模板参数模板型模板参数
这是《深入实践C++模板编程》第三章“模板参数类型详解”的读书笔记。
模板参数类型
模板参数通常表示类型，表示类型的模板参数称为类型模板参数（type template parameter）;此外还有非类型模板参数（non-type template parameter），非类型模板参数包含以下四...]]></description><category></category></item><item><title><![CDATA[MXNet Data IO]]></title><link>https://blog.youkuaiyun.com/KangRoger/article/details/73718712</link><guid>https://blog.youkuaiyun.com/KangRoger/article/details/73718712</guid><author>KangRoger</author><pubDate>Sun, 25 Jun 2017 16:25:42 +0800</pubDate><description><![CDATA[MXNet Data IO设计]]></description><category></category></item><item><title><![CDATA[PS-Lite源码分析]]></title><link>https://blog.youkuaiyun.com/KangRoger/article/details/73307685</link><guid>https://blog.youkuaiyun.com/KangRoger/article/details/73307685</guid><author>KangRoger</author><pubDate>Thu, 15 Jun 2017 23:09:08 +0800</pubDate><description><![CDATA[Parameter Server的PS-Lite实现]]></description><category></category></item><item><title><![CDATA[C++中的字符串流]]></title><link>https://blog.youkuaiyun.com/KangRoger/article/details/72758426</link><guid>https://blog.youkuaiyun.com/KangRoger/article/details/72758426</guid><author>KangRoger</author><pubDate>Thu, 25 May 2017 22:23:49 +0800</pubDate><description><![CDATA[字符串流stream，用来做数据类型转换；和C以及C++11实现对比。]]></description><category></category></item><item><title><![CDATA[程序员应该了解的数字]]></title><link>https://blog.youkuaiyun.com/KangRoger/article/details/72669975</link><guid>https://blog.youkuaiyun.com/KangRoger/article/details/72669975</guid><author>KangRoger</author><pubDate>Tue, 23 May 2017 23:25:42 +0800</pubDate><description><![CDATA[程序访问不同设备所需时间]]></description><category></category></item><item><title><![CDATA[《Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition》笔记]]></title><link>https://blog.youkuaiyun.com/KangRoger/article/details/70833669</link><guid>https://blog.youkuaiyun.com/KangRoger/article/details/70833669</guid><author>KangRoger</author><pubDate>Wed, 26 Apr 2017 23:18:07 +0800</pubDate><description><![CDATA[CNN要求输入size固定，作者提出SPP layer，去除了CNN这一限制。]]></description><category></category></item><item><title><![CDATA[《Rich feature hierarchies for accurate object detection and semantic segmentation》笔记]]></title><link>https://blog.youkuaiyun.com/KangRoger/article/details/70665385</link><guid>https://blog.youkuaiyun.com/KangRoger/article/details/70665385</guid><author>KangRoger</author><pubDate>Mon, 24 Apr 2017 21:35:49 +0800</pubDate><description><![CDATA[R-CNN，通过CNN提取特征，之后训练SVM分类器分类、回归确定目标区域]]></description><category></category></item><item><title><![CDATA[《A Neural Algorithm of Artistic Style》笔记]]></title><link>https://blog.youkuaiyun.com/KangRoger/article/details/70162886</link><guid>https://blog.youkuaiyun.com/KangRoger/article/details/70162886</guid><author>KangRoger</author><pubDate>Thu, 13 Apr 2017 22:36:21 +0800</pubDate><description><![CDATA[图像内容和风格都是比较抽象的，这篇文章通过深度学习，提取了图像内容和风格，并将两者合并，生成新的图像。]]></description><category></category></item><item><title><![CDATA[《Xception: Deep Learning with Depthwise Separable Convolutios》笔记]]></title><link>https://blog.youkuaiyun.com/KangRoger/article/details/69929915</link><guid>https://blog.youkuaiyun.com/KangRoger/article/details/69929915</guid><author>KangRoger</author><pubDate>Sun, 09 Apr 2017 20:49:13 +0800</pubDate><description><![CDATA[继Inception V1-V4之后，又一方法改进其结构。通过解耦和处理cross-channel相关和spatial相关，得到一个Xception结构网络]]></description><category></category></item><item><title><![CDATA[《Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning》笔记]]></title><link>https://blog.youkuaiyun.com/KangRoger/article/details/69488642</link><guid>https://blog.youkuaiyun.com/KangRoger/article/details/69488642</guid><author>KangRoger</author><pubDate>Thu, 06 Apr 2017 23:37:33 +0800</pubDate><description><![CDATA[Inception和Residual结合：Inception-v4和Inception-ResNet]]></description><category></category></item><item><title><![CDATA[《Rethinking the Inception Architecture for Computer Vision》笔记]]></title><link>https://blog.youkuaiyun.com/KangRoger/article/details/69218625</link><guid>https://blog.youkuaiyun.com/KangRoger/article/details/69218625</guid><author>KangRoger</author><pubDate>Tue, 04 Apr 2017 23:11:42 +0800</pubDate><description><![CDATA[Inception-V2设计介绍]]></description><category></category></item><item><title><![CDATA[《Residual Networks Behave Like Ensembles of Relatively Shallow Networks》笔记]]></title><link>https://blog.youkuaiyun.com/KangRoger/article/details/68938680</link><guid>https://blog.youkuaiyun.com/KangRoger/article/details/68938680</guid><author>KangRoger</author><pubDate>Fri, 31 Mar 2017 23:16:05 +0800</pubDate><description><![CDATA[从Ensemble角度解释Residual Networks]]></description><category></category></item><item><title><![CDATA[梯度下降算法总结]]></title><link>https://blog.youkuaiyun.com/KangRoger/article/details/64622249</link><guid>https://blog.youkuaiyun.com/KangRoger/article/details/64622249</guid><author>KangRoger</author><pubDate>Thu, 30 Mar 2017 21:54:43 +0800</pubDate><description><![CDATA[梯度下降法：基本梯度下降法、Momentum梯度下降法、Nesterov Momentum梯度下降法、AdaGrad、RMSprop、AdaDelta、Adam]]></description><category></category></item><item><title><![CDATA[《Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift》笔记]]></title><link>https://blog.youkuaiyun.com/KangRoger/article/details/62091914</link><guid>https://blog.youkuaiyun.com/KangRoger/article/details/62091914</guid><author>KangRoger</author><pubDate>Tue, 14 Mar 2017 23:35:20 +0800</pubDate><description><![CDATA[batch normalization]]></description><category></category></item><item><title><![CDATA[《Understanding the difficulty of training deep feedforward neural networks》笔记]]></title><link>https://blog.youkuaiyun.com/KangRoger/article/details/61414426</link><guid>https://blog.youkuaiyun.com/KangRoger/article/details/61414426</guid><author>KangRoger</author><pubDate>Sat, 11 Mar 2017 11:13:51 +0800</pubDate><description><![CDATA[Sigmod为什么不适合深度学习，交叉熵代价函数和平方差代价函数，神经网络权重初始化]]></description><category></category></item><item><title><![CDATA[机器学习中的损失函数]]></title><link>https://blog.youkuaiyun.com/KangRoger/article/details/59200996</link><guid>https://blog.youkuaiyun.com/KangRoger/article/details/59200996</guid><author>KangRoger</author><pubDate>Wed, 01 Mar 2017 23:40:38 +0800</pubDate><description><![CDATA[平方损失函数
对数损失函数
Hinge损失函数]]></description><category></category></item><item><title><![CDATA[ILSVRC历届冠军论文笔记]]></title><link>https://blog.youkuaiyun.com/KangRoger/article/details/56522132</link><guid>https://blog.youkuaiyun.com/KangRoger/article/details/56522132</guid><author>KangRoger</author><pubDate>Wed, 22 Feb 2017 23:39:28 +0800</pubDate><description><![CDATA[记录ILSVRC比赛，几个经典网络结构笔记]]></description><category></category></item><item><title><![CDATA[cs231n-(9)迁移学习和Fine-tune网络]]></title><link>https://blog.youkuaiyun.com/KangRoger/article/details/55681469</link><guid>https://blog.youkuaiyun.com/KangRoger/article/details/55681469</guid><author>KangRoger</author><pubDate>Sat, 18 Feb 2017 21:55:09 +0800</pubDate><description><![CDATA[介绍适合fine-tune的场景，以及方法。]]></description><category></category></item></channel></rss>