<?xml version="1.0" encoding="utf-8" ?><rss version="2.0"><channel><title><![CDATA[Joyliness]]></title><description><![CDATA[]]></description><link>https://blog.youkuaiyun.com/Joyliness</link><language>zh-cn</language><generator>https://blog.youkuaiyun.com/</generator><copyright><![CDATA[Copyright &copy; Joyliness]]></copyright><item><title><![CDATA[Selenium环境配置]]></title><link>https://blog.youkuaiyun.com/Joyliness/article/details/78657991</link><guid>https://blog.youkuaiyun.com/Joyliness/article/details/78657991</guid><author>Joyliness</author><pubDate>Sun, 16 Jun 2019 13:29:25 +0800</pubDate><description><![CDATA[官网下载Selenium最新版本

目前更新至3.7.1版，本地存放路径为D:\R\library\Rwebdriver\，名称为selenium-server-standalone-[version.number].jar，即selenium-server-standalone-3.7.1 






运行Selenium需要搭建Java环境

JDK（Java SE Development ...]]></description><category></category></item><item><title><![CDATA[Python绘制饼状图/甜甜圈]]></title><link>https://blog.youkuaiyun.com/Joyliness/article/details/83932591</link><guid>https://blog.youkuaiyun.com/Joyliness/article/details/83932591</guid><author>Joyliness</author><pubDate>Thu, 15 Nov 2018 17:31:46 +0800</pubDate><description><![CDATA[Python Matplotlib.pyplot
Matplotlib库是一个面向对象的绘图库。绘图界面由pyplot模块提供。该模块提供了许多绘图函数，以下记录的是饼状图/甜甜圈图的相关参数和绘图过程，官方资料详见文末链接。
# 导入相关模块
import numpy as np
import matplotlib.pyplot as plt
# 使中文正常显示的参数设置
plt.rcParam...]]></description><category></category></item><item><title><![CDATA[多元正态分布的极大似然估计]]></title><link>https://blog.youkuaiyun.com/Joyliness/article/details/80097491</link><guid>https://blog.youkuaiyun.com/Joyliness/article/details/80097491</guid><author>Joyliness</author><pubDate>Wed, 20 Jun 2018 13:53:12 +0800</pubDate><description><![CDATA[多元正态分布的极大似然估计



1. 一元正态分布的密度函数

一元正态分布的密度函数表示为：



f(x)=1(2π)−−−−√σe−(x−μ)22σ2f(x)=1(2π)σe−(x−μ)22σ2f(x) = \frac{1}{\sqrt {(2 \pi)} \sigma} e^{- \frac{(x - \mu)^2}{2 \sigma^2}}其中，σ&amp;amp;amp;amp;amp;amp;amp;amp;gt;0σ&amp;amp;amp;amp;amp;amp;amp;amp;gt;0]]></description><category></category></item><item><title><![CDATA[HMM的参数学习问题]]></title><link>https://blog.youkuaiyun.com/Joyliness/article/details/79603950</link><guid>https://blog.youkuaiyun.com/Joyliness/article/details/79603950</guid><author>Joyliness</author><pubDate>Mon, 19 Mar 2018 16:41:37 +0800</pubDate><description><![CDATA[HMM的参数学习问题

HMM的参数学习问题有两种：


监督学习：给定观测序列O=(o1,...,oT)O=(o1,...,oT)O = (o_1,...,o_T)和对应的状态序列I=(i1,...,iT)I=(i1,...,iT)I = (i_1,...,i_T)，估计参数λ=(A,B,π)λ=(A,B,π)\lambda = (A,B,\pi)。
非监督学习：只给定观测序列O=(o1,......]]></description><category></category></item><item><title><![CDATA[HMM的概率计算问题]]></title><link>https://blog.youkuaiyun.com/Joyliness/article/details/79593485</link><guid>https://blog.youkuaiyun.com/Joyliness/article/details/79593485</guid><author>Joyliness</author><pubDate>Sun, 18 Mar 2018 01:07:30 +0800</pubDate><description><![CDATA[HMM的概率计算问题

HMM的概率计算问题是指，给定模型参数λ=(A,B,π)λ=(A,B,π)\lambda = (A,B,\pi) 和观测序列O=(o1,o2,...,oT)O=(o1,o2,...,oT)O = (o_1,o_2,...,o_T)，计算在模型λλ\lambda下，观测序列OOO出现的概率：P(O|λ)P(O|λ)P(O | \lambda)。





直接计算

按概率公...]]></description><category></category></item><item><title><![CDATA[HMM的基本概念]]></title><link>https://blog.youkuaiyun.com/Joyliness/article/details/79592856</link><guid>https://blog.youkuaiyun.com/Joyliness/article/details/79592856</guid><author>Joyliness</author><pubDate>Sat, 17 Mar 2018 16:28:51 +0800</pubDate><description><![CDATA[HMM的变量定义

两个集合


长度为NNN的隐状态集合Q={q1,q2,...,qN}Q={q1,q2,...,qN}Q = \{q_1,q_2,...,q_N\}
长度为MMM的观测值集合V={v1,v2,...,vM}V={v1,v2,...,vM}V = \{v_1,v_2,...,v_M\}


两个序列


长度为TTT的隐状态序列I={i1,i2,...,iT}I={i1,i2,....]]></description><category></category></item><item><title><![CDATA[EM算法及对GMM的参数估计（EM算法的R实现 vs R mclust包）]]></title><link>https://blog.youkuaiyun.com/Joyliness/article/details/79504693</link><guid>https://blog.youkuaiyun.com/Joyliness/article/details/79504693</guid><author>Joyliness</author><pubDate>Sat, 10 Mar 2018 01:21:25 +0800</pubDate><description><![CDATA[EM算法与高斯混合模型（GMM）

EM算法是一种迭代算法，1977年由Dempster等人总结提出，用于含有隐变量（hidden variable）的概率模型参数的极大似然估计，或极大后验概率估计。EM算法的每次迭代由2步组成：


E-step：求期望（expectation）
M-step：最大化（maximization）


所以该算法称为期望极大算法（expectation maxim...]]></description><category></category></item><item><title><![CDATA[【杂纪】从ROC曲线到AUC值，再到Mann–Whitney U统计量]]></title><link>https://blog.youkuaiyun.com/Joyliness/article/details/79156879</link><guid>https://blog.youkuaiyun.com/Joyliness/article/details/79156879</guid><author>Joyliness</author><pubDate>Fri, 16 Feb 2018 23:01:00 +0800</pubDate><description><![CDATA[统计检验中的两类错误

在进行假设检验时，分别提出原假设（Null Hypothesis）和备择假设（Alternative Hypothesis），检验结果可能出现的两类错误：


原假设实际上是正确的，而检验结果却拒绝原假设，称为第一类/第一型错误（Type I error）、弃真错误
原假设实际上是错误的，而检验结果却接受原假设，称为第二类/第二型错误（Type II error）、取伪错误...]]></description><category></category></item><item><title><![CDATA[怕是诸葛先生听了也要为之动容吧~]]></title><link>https://blog.youkuaiyun.com/Joyliness/article/details/79160240</link><guid>https://blog.youkuaiyun.com/Joyliness/article/details/79160240</guid><author>Joyliness</author><pubDate>Thu, 25 Jan 2018 13:15:25 +0800</pubDate><description><![CDATA[优酷：王洛勇英文朗诵《出师表》


臣亮言：先帝创业未半而中道崩殂，

Permit me liang to observe: the late emperor was taken from us before he could finish his life’s work, the restoration of the Han.

今天下三分，益州疲弊，此诚危急存亡之秋也。

Today, th]]></description><category></category></item><item><title><![CDATA[糗事百科（Python(正则表达式、BeautifulSoup、Lxml) & R(rvest)）]]></title><link>https://blog.youkuaiyun.com/Joyliness/article/details/79116192</link><guid>https://blog.youkuaiyun.com/Joyliness/article/details/79116192</guid><author>Joyliness</author><pubDate>Tue, 23 Jan 2018 20:57:36 +0800</pubDate><description><![CDATA[Python

Python中用于爬取静态网页的基本方法/模块有三种：正则表达式、BeautifulSoup和Lxml。本文使用这三种方法分别抓取同样的数据，并计算各自的耗时，从而对三者的性能做简单探讨。三种方法的特点大致如下：




  爬取方法
  性能
  使用难度
  安装难度



  正则表达式
  快
  困难
  简单（内置模块）


  BeautifulSoup
  慢
  ...]]></description><category></category></item><item><title><![CDATA[【正则表达式】斗破苍穹（Python & R）]]></title><link>https://blog.youkuaiyun.com/Joyliness/article/details/79116005</link><guid>https://blog.youkuaiyun.com/Joyliness/article/details/79116005</guid><author>Joyliness</author><pubDate>Sat, 20 Jan 2018 19:23:11 +0800</pubDate><description><![CDATA[Python
# 加载模块
import re
import time
import requests# 伪装报头
headers = {
   'User-Agent' : 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) \
   AppleWebKit/537.36 (KHTML, like Gecko) Chrome/63.0.3239.84 Safari]]></description><category></category></item><item><title><![CDATA[【CSS Selector】酷狗Top500（Python & R）]]></title><link>https://blog.youkuaiyun.com/Joyliness/article/details/79106107</link><guid>https://blog.youkuaiyun.com/Joyliness/article/details/79106107</guid><author>Joyliness</author><pubDate>Fri, 19 Jan 2018 14:41:56 +0800</pubDate><description><![CDATA[Python




# 加载模块
import pandas
import requests
from bs4 import BeautifulSoup

# 伪装报头
headers = {
   'User-Agent' : 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) \
   AppleWebKit/537.36 (KHTML, like Geck]]></description><category></category></item><item><title><![CDATA[【CSS Selector】小猪网短租房（Python & R）]]></title><link>https://blog.youkuaiyun.com/Joyliness/article/details/79079813</link><guid>https://blog.youkuaiyun.com/Joyliness/article/details/79079813</guid><author>Joyliness</author><pubDate>Thu, 18 Jan 2018 21:07:21 +0800</pubDate><description><![CDATA[Python




# 加载模块
import csv
import time
import codecs
import random
import requests
from bs4 import BeautifulSoup

# 伪装报头
headers = {
   'User-Agent' : 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) \]]></description><category></category></item><item><title><![CDATA[RSelenium包抓取网易云音乐歌词（iframe框架的处理）]]></title><link>https://blog.youkuaiyun.com/Joyliness/article/details/78907969</link><guid>https://blog.youkuaiyun.com/Joyliness/article/details/78907969</guid><author>Joyliness</author><pubDate>Wed, 27 Dec 2017 01:19:32 +0800</pubDate><description><![CDATA[思路



Created with Raphaël 2.1.2网易云音乐《中国嘻哈榜》Step1：遍历《中国嘻哈榜》页面，获得每期链接programlinkStep2：遍历所有期数链接，获得每期歌曲链接musiclinkStep3：遍历每首歌曲链接，获得歌词lyricEnd





页面准备

library(RSelenium)
library(rvest)
base  &amp;lt;- &quot;htt...]]></description><category></category></item><item><title><![CDATA[【函数】R语言System {base}函数简介]]></title><link>https://blog.youkuaiyun.com/Joyliness/article/details/78882150</link><guid>https://blog.youkuaiyun.com/Joyliness/article/details/78882150</guid><author>Joyliness</author><pubDate>Sat, 23 Dec 2017 21:00:23 +0800</pubDate><description><![CDATA[system {base}
描述函数system的作用是，调用由参数command指定的操作系统命令。
用法
system(command, intern = FALSE, 
       ignore.stdout = FALSE, ignore.stderr = FALSE, 
       wait = TRUE, input = NULL, show.output.on.console =]]></description><category></category></item><item><title><![CDATA[RSelenium包抓取链家网（下：数据存储与容错管理）]]></title><link>https://blog.youkuaiyun.com/Joyliness/article/details/78817758</link><guid>https://blog.youkuaiyun.com/Joyliness/article/details/78817758</guid><author>Joyliness</author><pubDate>Sat, 16 Dec 2017 22:07:13 +0800</pubDate><description><![CDATA[HouseinfoFunc function(link) {
    destianation &amp;quot;UTF-8&amp;quot;)
    # 获取第i条房屋链接的页面内容
    location     % html_nodes(&amp;quot;a.no_resblock_a&amp;quot;) %&amp;gt;% html_text()
    # 小区位置
    unit         % html_nodes(&amp;quot;.price span.uni]]></description><category></category></item><item><title><![CDATA[RSelenium包抓取链家网（上：模拟点击与页面抓取）]]></title><link>https://blog.youkuaiyun.com/Joyliness/article/details/78801297</link><guid>https://blog.youkuaiyun.com/Joyliness/article/details/78801297</guid><author>Joyliness</author><pubDate>Thu, 14 Dec 2017 19:21:19 +0800</pubDate><description><![CDATA[安装RSelenium包install.packages(&quot;RSelenium&quot;)
# 直接从CRAN下载RSelenium包启动Selenium服务器
在控制台输入java -jar D:\R\library\Rwebdriver\selenium-server-standalone-3.7.1.jar以启动Selenium服务器。保持打开状态，可配合plantomjs、Chrome或Firefo]]></description><category></category></item><item><title><![CDATA[【包】R语言rvest包简介]]></title><link>https://blog.youkuaiyun.com/Joyliness/article/details/78722317</link><guid>https://blog.youkuaiyun.com/Joyliness/article/details/78722317</guid><author>Joyliness</author><pubDate>Sat, 09 Dec 2017 10:58:52 +0800</pubDate><description><![CDATA[rvest




  参考资料：
  
  rvest包github项目主页 
  rvest参考手册 
  rvest包文档简易翻译]]></description><category></category></item><item><title><![CDATA[【包】R语言rdom包简介]]></title><link>https://blog.youkuaiyun.com/Joyliness/article/details/78695549</link><guid>https://blog.youkuaiyun.com/Joyliness/article/details/78695549</guid><author>Joyliness</author><pubDate>Tue, 05 Dec 2017 14:04:52 +0800</pubDate><description><![CDATA[1. 官网下载Selenium最新版本



目前更新至3.7.1版，本地存放路径为D:\R\library\Rwebdriver\，名称为selenium-server-standalone-[version.number].jar，即selenium-server-standalone-3.7.1







2. 运行Selenium需要搭建Java环境



JDK（Java SE De]]></description><category></category></item><item><title><![CDATA[RSelenium/Rwebdriver/rdom包抓取表格数据]]></title><link>https://blog.youkuaiyun.com/Joyliness/article/details/78665139</link><guid>https://blog.youkuaiyun.com/Joyliness/article/details/78665139</guid><author>Joyliness</author><pubDate>Sun, 03 Dec 2017 13:16:34 +0800</pubDate><description><![CDATA[安装RSelenium/Rwebdriver

install.packages(&amp;quot;RSelenium&amp;quot;)
#直接从CRAN下载RSelenium包
install.packages('devtools');library(devtools);install_github(repo='Rwebdriver',username='crubba')
#从github下载Rwebdriver包

RSe]]></description><category></category></item></channel></rss>