<?xml version="1.0" encoding="utf-8" ?><rss version="2.0"><channel><title><![CDATA[jjcoder的专栏]]></title><description><![CDATA[]]></description><link>https://blog.youkuaiyun.com/u010956473</link><language>zh-cn</language><generator>https://blog.youkuaiyun.com/</generator><copyright><![CDATA[Copyright &copy; u010956473]]></copyright><item><title><![CDATA[基于深度学习的目标检测技术演进：R-CNN、Fast R-CNN、Faster R-CNN]]></title><link>https://blog.youkuaiyun.com/u010956473/article/details/79951797</link><guid>https://blog.youkuaiyun.com/u010956473/article/details/79951797</guid><author>u010956473</author><pubDate>Sun, 15 Apr 2018 18:50:42 +0800</pubDate><description><![CDATA[


    
        

        
        
        
            object detection我的理解，就是在给定的图片中精确找到物体所在位置，并标注出物体的类别。object detection要解决的问题就是物体在哪里，是什么这整个流程的问题。然而，这个问题可不是那么容易解决的，物体的尺寸变化范围很大，摆放物体的角度，姿态不定，而且可以出现在...]]></description><category></category></item><item><title><![CDATA[Practical Lessons from Predicting Clicks on Ads at Facebook]]></title><link>https://blog.youkuaiyun.com/u010956473/article/details/79943849</link><guid>https://blog.youkuaiyun.com/u010956473/article/details/79943849</guid><author>u010956473</author><pubDate>Sat, 14 Apr 2018 20:27:14 +0800</pubDate><description><![CDATA[Abstract这篇论文介绍 GBDT+LR 模型表现比单个模型好3%，另外本文探讨了一系列影响最终预测表现的因素。毫无疑问其中最重要的是 拥有正确的特征（捕捉用户和广告历史信息）+Model（GBDT+LR）的贡献程度最大 
数据实时性，学习速率，数据采样等其他因素影响较小。INTRODUCTION介绍了先前的一些相关paper。包括Google，Yahoo，MS的关于CTR Model方面的pa]]></description><category></category></item><item><title><![CDATA[深度剖析贝叶斯分类器]]></title><link>https://blog.youkuaiyun.com/u010956473/article/details/79903443</link><guid>https://blog.youkuaiyun.com/u010956473/article/details/79903443</guid><author>u010956473</author><pubDate>Wed, 11 Apr 2018 21:29:14 +0800</pubDate><description><![CDATA[

条件概率和全概率
全概率公式
客观概率 与主观概率


贝叶斯分析
先验概率
后验概率
极大似然估计


朴素贝叶斯分类器思路与方法
朴素贝叶斯的原理
贝叶斯极大似然参数估计
贝叶斯估计
贝叶斯应用







条件概率和全概率

条件概率描述的是事件A在另一事件B已发生的条件下的概率，记作P(A|B):

P(A|B)=P(A,B)P(B)P(A|B)=P(A,B)P(B)P(A|B) =...]]></description><category></category></item><item><title><![CDATA[基于深度学习的目标检测与识别]]></title><link>https://blog.youkuaiyun.com/u010956473/article/details/79206726</link><guid>https://blog.youkuaiyun.com/u010956473/article/details/79206726</guid><author>u010956473</author><pubDate>Tue, 30 Jan 2018 15:35:45 +0800</pubDate><description><![CDATA[FAIR的目标检测研究平台，实现Mask R-CNN，RetinaNet等流行算法。

Detectron是Facebook AI Research的软件系统，实现了最先进的目标检测算法，包括Mask R-CNN。 它是用Python编写的，由Caffe2深度学习框架支持。

在FAIR，Detectron已经启动了许多研究项目，包括：用于物体检测的特征金字塔网络，掩模R-CNN，检测和识别人 -]]></description><category></category></item><item><title><![CDATA[RCNN方法汇总及改进]]></title><link>https://blog.youkuaiyun.com/u010956473/article/details/79192375</link><guid>https://blog.youkuaiyun.com/u010956473/article/details/79192375</guid><author>u010956473</author><pubDate>Mon, 29 Jan 2018 10:30:02 +0800</pubDate><description><![CDATA[Faster R-CNN 从2015年底至今已经有接近两年了，但依旧还是Object Detection领域的主流框架之一，虽然推出了后续 R-FCN，Mask R-CNN 等改进框架，但基本结构变化不大。同时不乏有SSD，YOLO等骨骼清奇的新作，但精度上依然以Faster R-CNN为最好。对于一般的通用检测问题（例如行人检测，车辆检测，文字检测），只需在ImageNet
 pre-train]]></description><category></category></item><item><title><![CDATA[支持向量机（SVM）]]></title><link>https://blog.youkuaiyun.com/u010956473/article/details/78166707</link><guid>https://blog.youkuaiyun.com/u010956473/article/details/78166707</guid><author>u010956473</author><pubDate>Fri, 06 Oct 2017 19:53:58 +0800</pubDate><description><![CDATA[SVM定义是在特征空间上的间隔最大的线性分类器，间隔最大使它区别于感知机； 
有的人认为支持向量机是最好的现成分类器，SVM能够对训练集以外的数据点做出很好的决策。先介绍几个概念： 
一、数据的可分性： 
（1）当数据线性可分时，可以学习一个线性可分向量机 
（2）当数据不线性可分，可以利用核函数学习一个非线性分类器 
二、函数间隔 和几何间隔： 
A,B,C三个点均在正类一侧，A距离超平面较远，若]]></description><category></category></item><item><title><![CDATA[Numpy基础操作]]></title><link>https://blog.youkuaiyun.com/u010956473/article/details/78145836</link><guid>https://blog.youkuaiyun.com/u010956473/article/details/78145836</guid><author>u010956473</author><pubDate>Sat, 30 Sep 2017 23:18:11 +0800</pubDate><description><![CDATA[NumPy’s main object is the homogeneous multidimensional array. It is a table of elements (usually numbers), all of the same type, indexed by a tuple of positive integers. In NumPy dimensions are called]]></description><category></category></item><item><title><![CDATA[575. Distribute Candies]]></title><link>https://blog.youkuaiyun.com/u010956473/article/details/77129120</link><guid>https://blog.youkuaiyun.com/u010956473/article/details/77129120</guid><author>u010956473</author><pubDate>Sat, 12 Aug 2017 20:04:56 +0800</pubDate><description><![CDATA[一、问题：Given an integer array with even length, where different numbers in this array represent different kinds of candies. Each number means one candy of the corresponding kind. You need to distribute t]]></description><category></category></item><item><title><![CDATA[500. Keyboard Row]]></title><link>https://blog.youkuaiyun.com/u010956473/article/details/77127031</link><guid>https://blog.youkuaiyun.com/u010956473/article/details/77127031</guid><author>u010956473</author><pubDate>Sat, 12 Aug 2017 19:03:00 +0800</pubDate><description><![CDATA[一、题目：Given a List of words, return the words that can be typed using letters of alphabet on only one row's of American keyboard like the image below.二、解题class Solution {
    /*
    从第一个单词开始找，先找出该单词首字母在]]></description><category></category></item><item><title><![CDATA[STL容器   set、 multiset、map、multimap]]></title><link>https://blog.youkuaiyun.com/u010956473/article/details/77113015</link><guid>https://blog.youkuaiyun.com/u010956473/article/details/77113015</guid><author>u010956473</author><pubDate>Sat, 12 Aug 2017 11:08:48 +0800</pubDate><description><![CDATA[Set和Multiset 会根据特定的排序准则，自动将元素排序。两者不同之处是multiset允许元素重复而set不允许。set、 multiset的能力set和multiset通常以平衡二叉树完成，自动排序的优点在于令二叉树查找元素时拥有对数的性能。 
另外自动排序造成set、multiset有个重要的限制：不能改变元素值。set、 multiset的操作函数c.count(val);//返回元素]]></description><category></category></item><item><title><![CDATA[C++ STL 之 unordered_set 介绍]]></title><link>https://blog.youkuaiyun.com/u010956473/article/details/77100750</link><guid>https://blog.youkuaiyun.com/u010956473/article/details/77100750</guid><author>u010956473</author><pubDate>Fri, 11 Aug 2017 17:14:17 +0800</pubDate><description><![CDATA[简介          C++ 11中出现了两种新的关联容器:unordered_set和unordered_map，其内部实现与set和map大有不同，set和map内部实现是基于RB-Tree，而unordered_set和unordered_map内部实现是基于哈希表(hashtable)，由于unordered_set和unordered_map内部实现的公共接口大致相同，所以本文以unor]]></description><category></category></item><item><title><![CDATA[并发服务器的实现]]></title><link>https://blog.youkuaiyun.com/u010956473/article/details/76889475</link><guid>https://blog.youkuaiyun.com/u010956473/article/details/76889475</guid><author>u010956473</author><pubDate>Tue, 08 Aug 2017 11:18:44 +0800</pubDate><description><![CDATA[一、服务器端：tcpserv01.c/*************************************************************************
    > File Name: tcpserv01.c
    > Author: zsj
    > Mail: 15623582731.com 
    > Created Time: 2017年08月07日]]></description><category></category></item><item><title><![CDATA[并发服务器实现原理]]></title><link>https://blog.youkuaiyun.com/u010956473/article/details/76855767</link><guid>https://blog.youkuaiyun.com/u010956473/article/details/76855767</guid><author>u010956473</author><pubDate>Mon, 07 Aug 2017 19:29:04 +0800</pubDate><description><![CDATA[一、监听套接字与连接套接字的区别：并发服务器中主服务器通过派生一个子进程来处理每个新的连接。如果一个子进程继续使用服务器众所周知的端口来服务一个长时间的请求。首先服务器IP地址为 12.106.32.254和192.168.42.1.服务器在它端口上执行被动打开。这是个监听套接字对，服务器在本地接口的21端口等待连接请求。、IP地址为206.168.112.219的主机启动第一个客户，它对服务器执行]]></description><category></category></item><item><title><![CDATA[121. Best Time to Buy and Sell Stock]]></title><link>https://blog.youkuaiyun.com/u010956473/article/details/76845451</link><guid>https://blog.youkuaiyun.com/u010956473/article/details/76845451</guid><author>u010956473</author><pubDate>Mon, 07 Aug 2017 15:02:06 +0800</pubDate><description><![CDATA[一、问题：Say you have an array for which the ith element is the price of a given stock on day i.If you were only permitted to complete at most one transaction (ie, buy one and sell one share of the stock),]]></description><category></category></item><item><title><![CDATA[448、Find All Numbers Disappeared in an Array]]></title><link>https://blog.youkuaiyun.com/u010956473/article/details/76804503</link><guid>https://blog.youkuaiyun.com/u010956473/article/details/76804503</guid><author>u010956473</author><pubDate>Mon, 07 Aug 2017 09:27:18 +0800</pubDate><description><![CDATA[一、问题： Given an array nums, write a function to move all 0's to the end of it while maintaining the relative order of the non-zero elements.For example, given nums = [0, 1, 0, 3, 12], after calling your]]></description><category></category></item><item><title><![CDATA[561. Array Partition I]]></title><link>https://blog.youkuaiyun.com/u010956473/article/details/76789471</link><guid>https://blog.youkuaiyun.com/u010956473/article/details/76789471</guid><author>u010956473</author><pubDate>Sun, 06 Aug 2017 19:12:13 +0800</pubDate><description><![CDATA[一、问题： Given an array of 2n integers, your task is to group these integers into n pairs of integer, say (a1, b1), (a2, b2), ..., (an, bn) which makes sum of min(ai, bi) for all i from 1 to n as large as]]></description><category></category></item><item><title><![CDATA[27. Remove Element]]></title><link>https://blog.youkuaiyun.com/u010956473/article/details/76788655</link><guid>https://blog.youkuaiyun.com/u010956473/article/details/76788655</guid><author>u010956473</author><pubDate>Sun, 06 Aug 2017 19:03:13 +0800</pubDate><description><![CDATA[一、问题：Given an array and a value, remove all instances of that value in place and return the new length.Do not allocate extra space for another array, you must do this in place with constant memory.The]]></description><category></category></item><item><title><![CDATA[基本套接字函数]]></title><link>https://blog.youkuaiyun.com/u010956473/article/details/76782141</link><guid>https://blog.youkuaiyun.com/u010956473/article/details/76782141</guid><author>u010956473</author><pubDate>Sun, 06 Aug 2017 17:11:28 +0800</pubDate><description><![CDATA[一、socket函数为了执行网络I/O，一个进程必须做的第一件事就是调用socket函数，指定期望的通信协议类型。int socket(int family,int type,int protocol);//成功则返回正数，出错返回-1sockfd=socket(AF_INET,SOCK_STREAM,0)参数faminly :AF_INET(IPV4协议)，AF_INET6(IPV6协议)
typ]]></description><category></category></item><item><title><![CDATA[Socket套接字API简介]]></title><link>https://blog.youkuaiyun.com/u010956473/article/details/76773022</link><guid>https://blog.youkuaiyun.com/u010956473/article/details/76773022</guid><author>u010956473</author><pubDate>Sun, 06 Aug 2017 15:31:54 +0800</pubDate><description><![CDATA[一、套接字地址结构：struct in_addr 
{
in_addr_t   s_addr;  //32位IP地址，网络字节排序。
};struct sockaddr_in
{
  unint8_t      sin_len  //结构长度 无符号的八位整数
  sa_faminly_t  sin_family;//协议族 AF_INET ，是一个无符号字段表示套接字结构中的地址族。
  in_p]]></description><category></category></item><item><title><![CDATA[简单的时间服务器程序]]></title><link>https://blog.youkuaiyun.com/u010956473/article/details/76768764</link><guid>https://blog.youkuaiyun.com/u010956473/article/details/76768764</guid><author>u010956473</author><pubDate>Sun, 06 Aug 2017 13:59:20 +0800</pubDate><description><![CDATA[一、客户程序：/*************************************************************************
    > File Name: timecli.c
    > Author:zsj
    > Mail: 15623582731@163.com 
    > Created Time: 2017年08月06日 星期日 09时25分]]></description><category></category></item></channel></rss>