<?xml version="1.0" encoding="utf-8" ?><rss version="2.0"><channel><title><![CDATA[SmartDemo的博客]]></title><description><![CDATA[]]></description><link>https://blog.youkuaiyun.com/SmartDemo</link><language>zh-cn</language><generator>https://blog.youkuaiyun.com/</generator><copyright><![CDATA[Copyright &copy; SmartDemo]]></copyright><item><title><![CDATA[Image和base64相互转化]]></title><link>https://blog.youkuaiyun.com/SmartDemo/article/details/145378442</link><guid>https://blog.youkuaiyun.com/SmartDemo/article/details/145378442</guid><author>SmartDemo</author><pubDate>Mon, 27 Jan 2025 11:20:37 +0800</pubDate><description><![CDATA[【代码】Image和base64相互转化。]]></description><category></category></item><item><title><![CDATA[因果注意力（or 掩码注意力）代码实现]]></title><link>https://blog.youkuaiyun.com/SmartDemo/article/details/145263785</link><guid>https://blog.youkuaiyun.com/SmartDemo/article/details/145263785</guid><author>SmartDemo</author><pubDate>Mon, 20 Jan 2025 15:29:51 +0800</pubDate><description><![CDATA[【代码】因果注意力（or掩码注意力）代码实现。]]></description><category></category></item><item><title><![CDATA[kmeans实现]]></title><link>https://blog.youkuaiyun.com/SmartDemo/article/details/144672347</link><guid>https://blog.youkuaiyun.com/SmartDemo/article/details/144672347</guid><author>SmartDemo</author><pubDate>Mon, 23 Dec 2024 16:44:06 +0800</pubDate><description><![CDATA[【代码】kmeans实现。]]></description><category></category></item><item><title><![CDATA[Image\Numpy\Tensor互相转化]]></title><link>https://blog.youkuaiyun.com/SmartDemo/article/details/143211223</link><guid>https://blog.youkuaiyun.com/SmartDemo/article/details/143211223</guid><author>SmartDemo</author><pubDate>Thu, 24 Oct 2024 15:51:27 +0800</pubDate><description><![CDATA[【代码】Image\Numpy\Tensor互相转化。]]></description><category></category></item><item><title><![CDATA[datetime]]></title><link>https://blog.youkuaiyun.com/SmartDemo/article/details/142652958</link><guid>https://blog.youkuaiyun.com/SmartDemo/article/details/142652958</guid><author>SmartDemo</author><pubDate>Mon, 30 Sep 2024 11:24:00 +0800</pubDate><description><![CDATA[【代码】datetime。]]></description><category></category></item><item><title><![CDATA[访问生成器yield对象]]></title><link>https://blog.youkuaiyun.com/SmartDemo/article/details/141058760</link><guid>https://blog.youkuaiyun.com/SmartDemo/article/details/141058760</guid><author>SmartDemo</author><pubDate>Fri, 09 Aug 2024 12:21:36 +0800</pubDate><description><![CDATA[二、访问生成器对象

1、将生成器转换为列表


2、使用  函数来获取生成器的值


3、使用循环遍历生成器]]></description><category></category></item><item><title><![CDATA[从网络上下载并展示图像数据]]></title><link>https://blog.youkuaiyun.com/SmartDemo/article/details/141030231</link><guid>https://blog.youkuaiyun.com/SmartDemo/article/details/141030231</guid><author>SmartDemo</author><pubDate>Thu, 08 Aug 2024 16:24:22 +0800</pubDate><description><![CDATA[【代码】从网络上下载并展示图像数据。]]></description><category></category></item><item><title><![CDATA[logging日志实操入门]]></title><link>https://blog.youkuaiyun.com/SmartDemo/article/details/140997187</link><guid>https://blog.youkuaiyun.com/SmartDemo/article/details/140997187</guid><author>SmartDemo</author><pubDate>Wed, 07 Aug 2024 17:33:55 +0800</pubDate><description><![CDATA[logging日志实操入门]]></description><category></category></item><item><title><![CDATA[人工手动筛选特征]]></title><link>https://blog.youkuaiyun.com/SmartDemo/article/details/140871602</link><guid>https://blog.youkuaiyun.com/SmartDemo/article/details/140871602</guid><author>SmartDemo</author><pubDate>Fri, 02 Aug 2024 17:23:57 +0800</pubDate><description><![CDATA[提取和展示模型中各特征的重要性，尤其是查看特征重要性为零的特征以及非零特征。计算特征之间的相关性，并移除高度相关的特征（相关系数大于 0.95）使用信息值（IV）和群体稳定指数（PSI）筛选特征。]]></description><category></category></item><item><title><![CDATA[高度偏斜特征处理：log(x)、sqrt(x)、box-cox、Yeo-Johnson]]></title><link>https://blog.youkuaiyun.com/SmartDemo/article/details/140798152</link><guid>https://blog.youkuaiyun.com/SmartDemo/article/details/140798152</guid><author>SmartDemo</author><pubDate>Tue, 30 Jul 2024 15:36:34 +0800</pubDate><description><![CDATA[高度偏斜的特征 ： 数据分布不均匀、不对称的特征处理之后：使其分布更接近正态分布或至少减少偏斜程度。]]></description><category></category></item><item><title><![CDATA[python绘制单列堆积柱形图]]></title><link>https://blog.youkuaiyun.com/SmartDemo/article/details/136739004</link><guid>https://blog.youkuaiyun.com/SmartDemo/article/details/136739004</guid><author>SmartDemo</author><pubDate>Fri, 15 Mar 2024 14:30:00 +0800</pubDate><description><![CDATA[【代码】python绘制单列堆积柱形图。]]></description><category></category></item><item><title><![CDATA[python绘制双列堆积柱形图]]></title><link>https://blog.youkuaiyun.com/SmartDemo/article/details/136738703</link><guid>https://blog.youkuaiyun.com/SmartDemo/article/details/136738703</guid><author>SmartDemo</author><pubDate>Fri, 15 Mar 2024 14:22:29 +0800</pubDate><description><![CDATA[【代码】python绘制双列堆积柱形图。]]></description><category></category></item><item><title><![CDATA[google colab上解决matplotlib乱码]]></title><link>https://blog.youkuaiyun.com/SmartDemo/article/details/136738645</link><guid>https://blog.youkuaiyun.com/SmartDemo/article/details/136738645</guid><author>SmartDemo</author><pubDate>Fri, 15 Mar 2024 14:20:19 +0800</pubDate><description><![CDATA[【代码】google colab上解决matplotlib乱码。]]></description><category></category></item><item><title><![CDATA[DDPM单步的具体步骤]]></title><link>https://blog.youkuaiyun.com/SmartDemo/article/details/135489815</link><guid>https://blog.youkuaiyun.com/SmartDemo/article/details/135489815</guid><author>SmartDemo</author><pubDate>Tue, 09 Jan 2024 21:31:43 +0800</pubDate><description><![CDATA[在前向过程中，模型逐步向数据添加噪声，直到数据完全转化为无结构的噪声这个过程可以用马尔可夫链来描述，其中每一步都向数据添加一小部分高斯噪声。]]></description><category></category></item><item><title><![CDATA[DDPM之反向传播]]></title><link>https://blog.youkuaiyun.com/SmartDemo/article/details/135488005</link><guid>https://blog.youkuaiyun.com/SmartDemo/article/details/135488005</guid><author>SmartDemo</author><pubDate>Tue, 09 Jan 2024 19:26:14 +0800</pubDate><description><![CDATA[是由模型学习得到的，目的是最大程度地还原出原始数据。]]></description><category></category></item><item><title><![CDATA[DDPM之前向扩散]]></title><link>https://blog.youkuaiyun.com/SmartDemo/article/details/135480737</link><guid>https://blog.youkuaiyun.com/SmartDemo/article/details/135480737</guid><author>SmartDemo</author><pubDate>Tue, 09 Jan 2024 15:17:19 +0800</pubDate><description><![CDATA[前向过程。]]></description><category></category></item><item><title><![CDATA[扩散模型奠基之作：DDPM]]></title><link>https://blog.youkuaiyun.com/SmartDemo/article/details/135480523</link><guid>https://blog.youkuaiyun.com/SmartDemo/article/details/135480523</guid><author>SmartDemo</author><pubDate>Tue, 09 Jan 2024 14:55:54 +0800</pubDate><description><![CDATA[DDPM分为两个部分，设计两个不同的概率分布1、前向过程（扩散）：这一过程可以用条件概率 q(xt​∣xt−1​) 来描述2、反向过程（去噪）：这一过程可以用条件概率 p(xt−1​∣xt​) 来描述。]]></description><category></category></item><item><title><![CDATA[解决matplotlib中文乱码问题]]></title><link>https://blog.youkuaiyun.com/SmartDemo/article/details/135452039</link><guid>https://blog.youkuaiyun.com/SmartDemo/article/details/135452039</guid><author>SmartDemo</author><pubDate>Mon, 08 Jan 2024 11:07:30 +0800</pubDate><description><![CDATA[重点关注matplotlib中的ttf文件夹和matplotlibrc文档。第一步，将下载好的SimHei.ttf文件拷贝到ttf文件夹；提醒一点，不要忘记去掉上述所修改三条配置命令前的 # 符号。如果 _rebuild() 报错，则换一种方法reload。]]></description><category></category></item><item><title><![CDATA[核方法 ： 多项式核函数]]></title><link>https://blog.youkuaiyun.com/SmartDemo/article/details/135170211</link><guid>https://blog.youkuaiyun.com/SmartDemo/article/details/135170211</guid><author>SmartDemo</author><pubDate>Sat, 23 Dec 2023 16:11:35 +0800</pubDate><description><![CDATA[多项式核函数： 将数据映射到高维空间，从而实现低维线性不可分 到 高维线性可分]]></description><category></category></item><item><title><![CDATA[对图片进行crop操作]]></title><link>https://blog.youkuaiyun.com/SmartDemo/article/details/135158826</link><guid>https://blog.youkuaiyun.com/SmartDemo/article/details/135158826</guid><author>SmartDemo</author><pubDate>Fri, 22 Dec 2023 18:02:30 +0800</pubDate><description><![CDATA[【代码】对图片进行crop操作。]]></description><category></category></item></channel></rss>