<?xml version="1.0" encoding="utf-8" ?><rss version="2.0"><channel><title><![CDATA[enockyang的博客]]></title><description><![CDATA[]]></description><link>https://blog.youkuaiyun.com/enockyang</link><language>zh-cn</language><generator>https://blog.youkuaiyun.com/</generator><copyright><![CDATA[Copyright &copy; enockyang]]></copyright><item><title><![CDATA[minimum spanning tree]]></title><link>https://blog.youkuaiyun.com/enockyang/article/details/79946366</link><guid>https://blog.youkuaiyun.com/enockyang/article/details/79946366</guid><author>enockyang</author><pubDate>Sun, 15 Apr 2018 11:53:42 +0800</pubDate><description><![CDATA[首先我们要给出下面一个概念，如何可以节约空间地保存无向图：    a = [100 for i in range(int(10*11/2))]
    def update(i,j,l):
        if i &amp;gt; j:
            a[int(i*(i+1)/2+j)] = l
        else:
            a[int(j*(j+1)/2+i)] = ...]]></description><category></category></item><item><title><![CDATA[shortest path 实现]]></title><link>https://blog.youkuaiyun.com/enockyang/article/details/79922591</link><guid>https://blog.youkuaiyun.com/enockyang/article/details/79922591</guid><author>enockyang</author><pubDate>Fri, 13 Apr 2018 03:24:53 +0800</pubDate><description><![CDATA[使用python内置函数直接实现Dij算法，每次用全部已包含集中的点进行计算：def Dij(start):
    list_a = [1,2,3,4,5,6,7,8,9,10] ##所有点的集合
    list_s = [start]  ##最短路径已经确定的点的集合s
    dist = {} ##最后的结果，存为一个dictionary，key是点的数字，value是从start出发最...]]></description><category></category></item><item><title><![CDATA[杂记]]></title><link>https://blog.youkuaiyun.com/enockyang/article/details/79823646</link><guid>https://blog.youkuaiyun.com/enockyang/article/details/79823646</guid><author>enockyang</author><pubDate>Thu, 05 Apr 2018 03:37:46 +0800</pubDate><description><![CDATA[tuningGBM/随机森林调参 forloop --&amp;gt; use the best model]]></description><category></category></item><item><title><![CDATA[机器学习-决策树]]></title><link>https://blog.youkuaiyun.com/enockyang/article/details/79814877</link><guid>https://blog.youkuaiyun.com/enockyang/article/details/79814877</guid><author>enockyang</author><pubDate>Wed, 04 Apr 2018 11:59:14 +0800</pubDate><description><![CDATA[笔记参考：https://blog.youkuaiyun.com/GitChat/article/details/78967193顾名思义，决策树就是用一棵树来表示我们的整个决策过程，有点像人类做决策时的判断方式。这棵树可以是二叉树（比如 CART 只能是二叉树），也可以是多叉树（比如 ID3、C4.5 可以是多叉树或二叉树）。根节点包含整个样本集，每个叶节都对应一个决策结果（注意，不同的叶节点可能对应同一个...]]></description><category></category></item><item><title><![CDATA[机器学习知识]]></title><link>https://blog.youkuaiyun.com/enockyang/article/details/79797550</link><guid>https://blog.youkuaiyun.com/enockyang/article/details/79797550</guid><author>enockyang</author><pubDate>Tue, 03 Apr 2018 03:27:44 +0800</pubDate><description><![CDATA[https://blog.youkuaiyun.com/sinat_34496643/article/details/79035830]]></description><category></category></item><item><title><![CDATA[最长路径问题]]></title><link>https://blog.youkuaiyun.com/enockyang/article/details/79748657</link><guid>https://blog.youkuaiyun.com/enockyang/article/details/79748657</guid><author>enockyang</author><pubDate>Thu, 29 Mar 2018 20:37:17 +0800</pubDate><description><![CDATA[]]></description><category></category></item><item><title><![CDATA[在线编程例题练习， HR姐姐看人]]></title><link>https://blog.youkuaiyun.com/enockyang/article/details/79721986</link><guid>https://blog.youkuaiyun.com/enockyang/article/details/79721986</guid><author>enockyang</author><pubDate>Wed, 28 Mar 2018 08:41:53 +0800</pubDate><description><![CDATA[可以发现我们可以用一个递推公式来计算， S(N=4) = S(N=3) +new_pointsnew_points是新加的点中没有被原来点挡住的点第一步：仔细想来，如何判断一个点会不会被原来的点挡住 --&amp;gt; 相当于判断这个点的两个坐标是不是互质的点(8,4)很明显会被点(4,2)，而点(8,7)和点(8,1)就不会被之前的点挡住第二步：其实新增的点只需要算一半就可以了，因为图形和点是关于x=...]]></description><category></category></item><item><title><![CDATA[线性回归和逻辑回归 记录]]></title><link>https://blog.youkuaiyun.com/enockyang/article/details/79721790</link><guid>https://blog.youkuaiyun.com/enockyang/article/details/79721790</guid><author>enockyang</author><pubDate>Wed, 28 Mar 2018 04:21:57 +0800</pubDate><description><![CDATA[1 线性回归相关概念1.1 R-squareR-square值表示现有模型可以解释的方差占总体方差的比例SSTO = SSE +SSRR-square = SSR/SST0希望其接近11.2 F-value题外话：    t检验 --&amp;gt; 检验的对象是mean    f检验 --&amp;gt; 检验的对象是varianceF-value = MSR/MSE --&amp;gt; 希望其很大 --&amp;gt; 相...]]></description><category></category></item></channel></rss>