<?xml version="1.0" encoding="utf-8" ?><rss version="2.0"><channel><title><![CDATA[ABU366的博客]]></title><description><![CDATA[]]></description><link>https://blog.youkuaiyun.com/ABU366</link><language>zh-cn</language><generator>https://blog.youkuaiyun.com/</generator><copyright><![CDATA[Copyright &copy; ABU366]]></copyright><item><title><![CDATA[神经网络与深度学习第四章前馈神经网络习题解答]]></title><link>https://blog.youkuaiyun.com/ABU366/article/details/134072389</link><guid>https://blog.youkuaiyun.com/ABU366/article/details/134072389</guid><author>ABU366</author><pubDate>Fri, 27 Oct 2023 11:28:11 +0800</pubDate><description><![CDATA[再举个形象的例子，比如你要从山顶走到山谷，你直线走下去即为最大负梯度方向，收敛快，但是你要很贪心的曲折的走，虽然在朝着山谷的方向走，但不是最快的方向，你同样可以到达终点，但是你的速度相较于直线走肯定是要慢的，条条大路通罗马，但是不可能全部人都走最快的大路，也有人走蜿蜒曲折的小路。肯定不是沿着最大的负梯度方向进行下降的，而是以曲折抖动的方式进行下降（结合图1），但是以零均值化的方式，可以最大程度的保证数据沿着最大的负梯度方向进行下降，所以当。是恒大于0的，其次，若x恒大于0,会导致。]]></description><category></category></item><item><title><![CDATA[力扣--整数转罗马数字--Python]]></title><link>https://blog.youkuaiyun.com/ABU366/article/details/130541097</link><guid>https://blog.youkuaiyun.com/ABU366/article/details/130541097</guid><author>ABU366</author><pubDate>Sun, 07 May 2023 12:23:19 +0800</pubDate><description><![CDATA[2.如果num<1000,去哈希表中由高到低判断含有的最大值，并加n个s[i]字符。3.一直到num属于1-9的范围内，进行最后处理，跳出循环。1.如果num>=1000，则加n个'M'字符。]]></description><category></category></item><item><title><![CDATA[力扣--分情况实现N字形变换-python]]></title><link>https://blog.youkuaiyun.com/ABU366/article/details/130498961</link><guid>https://blog.youkuaiyun.com/ABU366/article/details/130498961</guid><author>ABU366</author><pubDate>Fri, 05 May 2023 08:45:11 +0800</pubDate><description><![CDATA[先处理第一行，根据第一行的索引处理中间的numrows-2行，最后处理最后一行。]]></description><category></category></item><item><title><![CDATA[力扣--哈希表实现无重复字符的最长子串]]></title><link>https://blog.youkuaiyun.com/ABU366/article/details/130483352</link><guid>https://blog.youkuaiyun.com/ABU366/article/details/130483352</guid><author>ABU366</author><pubDate>Thu, 04 May 2023 11:21:23 +0800</pubDate><description><![CDATA[3.对于正常输入，如果没有在哈希表中，先将max_length更新，以防丢失最大值，然后找到相同字符的索引，将该字符及之前的字符全部删除，添加最新的键值对，更新哈希表，将length变成哈希表的长度。2.对于正常输入，创建哈希表，如果没有在哈希表中，则添加，length++不仅可以计算出无重复最长子串的长度，也可以计算出最后的最长字串。4.对于一直进入if的输入，更新max_length，结束。添加print(index)查看中间状态。1.判断输入是否为空，若为空，返回0。]]></description><category></category></item><item><title><![CDATA[2022_3 CSP JPEG解码 Python]]></title><link>https://blog.youkuaiyun.com/ABU366/article/details/129290117</link><guid>https://blog.youkuaiyun.com/ABU366/article/details/129290117</guid><author>ABU366</author><pubDate>Wed, 01 Mar 2023 21:24:53 +0800</pubDate><description><![CDATA[CSP 2022_3 python JPEG解码]]></description><category></category></item><item><title><![CDATA[神经网络与深度学习作业12：第七章课后题]]></title><link>https://blog.youkuaiyun.com/ABU366/article/details/128263832</link><guid>https://blog.youkuaiyun.com/ABU366/article/details/128263832</guid><author>ABU366</author><pubDate>Sat, 10 Dec 2022 15:25:15 +0800</pubDate><description><![CDATA[NNDL 作业12：第七章课后题]]></description><category></category></item><item><title><![CDATA[神经网络与深度学习day16:网络优化与正则化（3）不同优化算法比较]]></title><link>https://blog.youkuaiyun.com/ABU366/article/details/128183910</link><guid>https://blog.youkuaiyun.com/ABU366/article/details/128183910</guid><author>ABU366</author><pubDate>Fri, 09 Dec 2022 21:47:27 +0800</pubDate><description><![CDATA[NNDL 实验八 网络优化与正则化（3）不同优化算法比较]]></description><category></category></item><item><title><![CDATA[NNDL 作业11：优化算法比较]]></title><link>https://blog.youkuaiyun.com/ABU366/article/details/128163444</link><guid>https://blog.youkuaiyun.com/ABU366/article/details/128163444</guid><author>ABU366</author><pubDate>Sat, 03 Dec 2022 20:57:21 +0800</pubDate><description><![CDATA[NNDL 作业11：优化算法比较]]></description><category></category></item><item><title><![CDATA[神经网络与深度学习day15:循环神经网络LSTM的记忆能力的实验]]></title><link>https://blog.youkuaiyun.com/ABU366/article/details/128081895</link><guid>https://blog.youkuaiyun.com/ABU366/article/details/128081895</guid><author>ABU366</author><pubDate>Thu, 01 Dec 2022 22:35:08 +0800</pubDate><description><![CDATA[NNDL 实验七 循环神经网络（3）LSTM的记忆能力实验]]></description><category></category></item><item><title><![CDATA[神经网络与深度学习作业10：（LSTM | GRU）]]></title><link>https://blog.youkuaiyun.com/ABU366/article/details/128052571</link><guid>https://blog.youkuaiyun.com/ABU366/article/details/128052571</guid><author>ABU366</author><pubDate>Sat, 26 Nov 2022 18:36:05 +0800</pubDate><description><![CDATA[神经网络与深度学习作业10：（LSTM | GRU）]]></description><category></category></item><item><title><![CDATA[神经网络与深度学习作业9：分别使用numpy和pytorch实现BPTT]]></title><link>https://blog.youkuaiyun.com/ABU366/article/details/128000198</link><guid>https://blog.youkuaiyun.com/ABU366/article/details/128000198</guid><author>ABU366</author><pubDate>Wed, 23 Nov 2022 15:19:58 +0800</pubDate><description><![CDATA[NNDL 作业9：分别使用numpy和pytorch实现BPTT]]></description><category></category></item><item><title><![CDATA[神经网络与深度学习实验day14-循环神经网络（2）梯度爆炸实验]]></title><link>https://blog.youkuaiyun.com/ABU366/article/details/127961612</link><guid>https://blog.youkuaiyun.com/ABU366/article/details/127961612</guid><author>ABU366</author><pubDate>Mon, 21 Nov 2022 13:36:38 +0800</pubDate><description><![CDATA[NNDL 实验七 循环神经网络（2）梯度爆炸实验]]></description><category></category></item><item><title><![CDATA[神经网络与深度学习实验day13-循环神经网络：RNN记忆能力实验]]></title><link>https://blog.youkuaiyun.com/ABU366/article/details/127912426</link><guid>https://blog.youkuaiyun.com/ABU366/article/details/127912426</guid><author>ABU366</author><pubDate>Thu, 17 Nov 2022 23:07:39 +0800</pubDate><description><![CDATA[NNDL 实验七 循环神经网络（1）RNN记忆能力实验]]></description><category></category></item><item><title><![CDATA[神经网络与深度学习作业8：RNN - 简单循环网络]]></title><link>https://blog.youkuaiyun.com/ABU366/article/details/127824295</link><guid>https://blog.youkuaiyun.com/ABU366/article/details/127824295</guid><author>ABU366</author><pubDate>Sat, 12 Nov 2022 23:07:56 +0800</pubDate><description><![CDATA[NNDL 作业8：RNN - 简单循环网络]]></description><category></category></item><item><title><![CDATA[神经网络与深度学习实验day12-基于ResNet18完成cifar-10图像分类任务]]></title><link>https://blog.youkuaiyun.com/ABU366/article/details/127772903</link><guid>https://blog.youkuaiyun.com/ABU366/article/details/127772903</guid><author>ABU366</author><pubDate>Wed, 09 Nov 2022 23:11:17 +0800</pubDate><description><![CDATA[神经网络与深度学习实验day12-基于ResNet18完成cifar-10图像分类任务]]></description><category></category></item><item><title><![CDATA[神经网络与深度学习实验day11-基于torch用ResNet18模型实现MNIST]]></title><link>https://blog.youkuaiyun.com/ABU366/article/details/127644902</link><guid>https://blog.youkuaiyun.com/ABU366/article/details/127644902</guid><author>ABU366</author><pubDate>Wed, 02 Nov 2022 20:58:04 +0800</pubDate><description><![CDATA[NNDL 实验六 卷积神经网络（4）ResNet18实现MNIST]]></description><category></category></item><item><title><![CDATA[神经网络与深度学习作业7-第五章课后题（1×1 卷积核 | CNN BP）]]></title><link>https://blog.youkuaiyun.com/ABU366/article/details/127586130</link><guid>https://blog.youkuaiyun.com/ABU366/article/details/127586130</guid><author>ABU366</author><pubDate>Sat, 29 Oct 2022 16:20:42 +0800</pubDate><description><![CDATA[NNDL 作业7：第五章课后题（1×1 卷积核 | CNN BP）]]></description><category></category></item><item><title><![CDATA[【神经网络与深度学习day10-基于torch使用Lenet实现手写数字识别】]]></title><link>https://blog.youkuaiyun.com/ABU366/article/details/127562955</link><guid>https://blog.youkuaiyun.com/ABU366/article/details/127562955</guid><author>ABU366</author><pubDate>Fri, 28 Oct 2022 22:10:09 +0800</pubDate><description><![CDATA[神经网络与深度学习day10：Lenet实现MNIST]]></description><category></category></item><item><title><![CDATA[pycharm和jupyter关于加载模型的区别]]></title><link>https://blog.youkuaiyun.com/ABU366/article/details/127475382</link><guid>https://blog.youkuaiyun.com/ABU366/article/details/127475382</guid><author>ABU366</author><pubDate>Sun, 23 Oct 2022 15:26:08 +0800</pubDate><description><![CDATA[pycharm和jupyter关于加载模型的区别]]></description><category></category></item><item><title><![CDATA[神经网络与深度学习作业6-XO识别]]></title><link>https://blog.youkuaiyun.com/ABU366/article/details/127453249</link><guid>https://blog.youkuaiyun.com/ABU366/article/details/127453249</guid><author>ABU366</author><pubDate>Sat, 22 Oct 2022 11:25:02 +0800</pubDate><description><![CDATA[神经网络与深度学习作业6-XO识别]]></description><category></category></item></channel></rss>