<?xml version="1.0" encoding="utf-8" ?><rss version="2.0"><channel><title><![CDATA[renxingkai的博客]]></title><description><![CDATA[]]></description><link>https://blog.youkuaiyun.com/renxingkai</link><language>zh-cn</language><generator>https://blog.youkuaiyun.com/</generator><copyright><![CDATA[Copyright &copy; renxingkai]]></copyright><item><title><![CDATA[机器阅读理解那些事儿]]></title><link>https://blog.youkuaiyun.com/renxingkai/article/details/88046325</link><guid>https://blog.youkuaiyun.com/renxingkai/article/details/88046325</guid><author>renxingkai</author><pubDate>Thu, 28 Feb 2019 22:14:11 +0800</pubDate><description><![CDATA[今日，“机器之心”公众号刷爆了超强小姐姐陈丹琦的文章，笔者前两天也刚看了小姐姐的论文，当时无意中点开开源的github链接，发现小姐姐一路绿灯的github也是十分敬佩！

再放一张小姐姐的主页图片，再次膜拜下~

小姐姐做的方向是机器阅读理解和智能问答(QA)，博主也看了一段时间的该方向论文，现在来聊一聊机器阅读理解中的世界~
阅读理解，其实我们大家从小到大都在做，从语文阅读理解到英语阅读理解，...]]></description><category></category></item><item><title><![CDATA[PyTorch可视化工具Visdom实战]]></title><link>https://blog.youkuaiyun.com/renxingkai/article/details/87904726</link><guid>https://blog.youkuaiyun.com/renxingkai/article/details/87904726</guid><author>renxingkai</author><pubDate>Sun, 24 Feb 2019 16:35:22 +0800</pubDate><description><![CDATA[Visdom是一个灵活的可视化工具，可以实时显示新创建的数据，支持Torch和Numpy，是深度学习框架PyTorch的可视化工具，做出的图像也较为美观，本文总结了Visdom的基础画图操作。
安装方式
在Python3环境下，使用pip直接安装Visdom模块。
pip install visdom

然后命令行启动服务器
python -m visdom.server

一旦启动了服务器后，就...]]></description><category></category></item><item><title><![CDATA[BERT_中文情感分类操作及代码]]></title><link>https://blog.youkuaiyun.com/renxingkai/article/details/87605693</link><guid>https://blog.youkuaiyun.com/renxingkai/article/details/87605693</guid><author>renxingkai</author><pubDate>Mon, 18 Feb 2019 11:04:49 +0800</pubDate><description><![CDATA[本实验，是用BERT进行中文情感分类，记录了详细操作及完整程序，代码链接，喜欢的话给个star哟(凑不要脸~)
本文参考奇点机智的文章，记录自己在运行BERT中的一些操作。
BERT的代码同论文里描述的一致，主要分为两个部分。一个是训练语言模型（language model）的预训练（pretrain）部分。另一个是训练具体任务(task)的fine-tune部分。
在开源的代码中，预训练的入口是...]]></description><category></category></item><item><title><![CDATA[神经网络结构可视化(Keras版)]]></title><link>https://blog.youkuaiyun.com/renxingkai/article/details/84871842</link><guid>https://blog.youkuaiyun.com/renxingkai/article/details/84871842</guid><author>renxingkai</author><pubDate>Fri, 07 Dec 2018 11:24:50 +0800</pubDate><description><![CDATA[mark一下:
from keras.utils.vis_utils import model_to_dot
from IPython.display import SVG
#model即为要可视化的网络模型
SVG(model_to_dot(model).create(prog='dot', format='svg'))

demo效果如下：


...]]></description><category></category></item><item><title><![CDATA[深度学习关键网络架构及其在Keras中的应用]]></title><link>https://blog.youkuaiyun.com/renxingkai/article/details/84258493</link><guid>https://blog.youkuaiyun.com/renxingkai/article/details/84258493</guid><author>renxingkai</author><pubDate>Mon, 19 Nov 2018 19:20:01 +0800</pubDate><description><![CDATA[最近撸完了《Python深度学习》，对深度学习又加深了认识，“深度学习真的是一门艺术”，以下是原书第9章的总结部分，很多heuristic知识都是前人总结下来的，个人觉得很有借鉴意义，因此mark下。
首先，我们来快速看一下输入模式与适当的网络架构之间的对应关系。

向量数据：密集连接网络（ Dense 层）。
图像数据：二维卷积神经网络。
声音数据（比如波形）：一维卷积神经网络（首选）或循环神经...]]></description><category></category></item><item><title><![CDATA[LeetCode-771-Jewels and Stones(宝石和石头)]]></title><link>https://blog.youkuaiyun.com/renxingkai/article/details/81591544</link><guid>https://blog.youkuaiyun.com/renxingkai/article/details/81591544</guid><author>renxingkai</author><pubDate>Sat, 11 Aug 2018 22:05:00 +0800</pubDate><description><![CDATA[Q：

You're given strings J representing the types of stones that are jewels, and S representing the stones you have.  Each character in S is a type of stone you have.  You want to know how many of the...]]></description><category></category></item><item><title><![CDATA[TensorFlow实现多层感知机]]></title><link>https://blog.youkuaiyun.com/renxingkai/article/details/81142290</link><guid>https://blog.youkuaiyun.com/renxingkai/article/details/81142290</guid><author>renxingkai</author><pubDate>Sat, 21 Jul 2018 10:57:02 +0800</pubDate><description><![CDATA[此代码主要参考黄文坚老师的《TensorFlow实战》

此网络数据集为MNIST手写数据集，采用一层隐含层（隐含层节点数量300）的神经网络架构，并且使用Dropout方法随机丢失神经元来丰富训练的特征，使用自适应学习速率算法Adagrad，激活函数使用ReLU，输出层采用softmax计算各个类别概率。

此处输入，[None,784]代表将图片784像素点一维“铺开”，一张图片对应一个向量：...]]></description><category></category></item><item><title><![CDATA[随机梯度下降法和最小二乘法的TensorFlow实现]]></title><link>https://blog.youkuaiyun.com/renxingkai/article/details/81098489</link><guid>https://blog.youkuaiyun.com/renxingkai/article/details/81098489</guid><author>renxingkai</author><pubDate>Wed, 18 Jul 2018 15:23:36 +0800</pubDate><description><![CDATA[1.随机梯度下降法（SGD）

随机梯度下降法是用来求参数的优化算法，具体算法不过多阐述，此处采用线性拟合来实现SGD，同时使用TensorFlow进行计算，具体思路注释写的较为清楚，直接附上代码：


import tensorflow as tf
import numpy as np
import matplotlib.pyplot as plt

#权重
W=0.3
#偏置
b=0.8

n...]]></description><category></category></item><item><title><![CDATA[随机梯度下降法学习]]></title><link>https://blog.youkuaiyun.com/renxingkai/article/details/80523281</link><guid>https://blog.youkuaiyun.com/renxingkai/article/details/80523281</guid><author>renxingkai</author><pubDate>Thu, 31 May 2018 14:00:35 +0800</pubDate><description><![CDATA[以下代码是学习李宏毅老师的课程，数据集为神奇宝贝进化后的cp数据值。# -*- coding: utf-8 -*-
&quot;&quot;&quot;
Created on Thu May 31 09:59:54 2018

@author: Administrator
&quot;&quot;&quot;

import numpy as np
import matplotlib.pyplot as plt

#给出训练数据
x_data=[328,3...]]></description><category></category></item><item><title><![CDATA[《scikit-learn常用机器学习算法》学习笔记---KNN算法及KNN回归]]></title><link>https://blog.youkuaiyun.com/renxingkai/article/details/80346577</link><guid>https://blog.youkuaiyun.com/renxingkai/article/details/80346577</guid><author>renxingkai</author><pubDate>Thu, 17 May 2018 10:04:09 +0800</pubDate><description><![CDATA[一、KNN算法具体算法原理不再细说，此处生成3类，60个点集合，并对预测值进行预测分类。# -*- coding: utf-8 -*-
&quot;&quot;&quot;
Created on Wed May 16 12:09:34 2018

@author: Administrator
&quot;&quot;&quot;

from sklearn.datasets.samples_generator import make_blobs
from ...]]></description><category></category></item><item><title><![CDATA[HDOJ-1019(Java)]]></title><link>https://blog.youkuaiyun.com/renxingkai/article/details/79759986</link><guid>https://blog.youkuaiyun.com/renxingkai/article/details/79759986</guid><author>renxingkai</author><pubDate>Fri, 30 Mar 2018 16:45:10 +0800</pubDate><description><![CDATA[Q:

The least common multiple (LCM) of a set of positive integers is the smallest positive integer which is divisible by all the numbers in the set. For example, the LCM of 5, 7 and 15 is 105.

Input:...]]></description><category></category></item><item><title><![CDATA[HDOJ-1005(Java)]]></title><link>https://blog.youkuaiyun.com/renxingkai/article/details/79743063</link><guid>https://blog.youkuaiyun.com/renxingkai/article/details/79743063</guid><author>renxingkai</author><pubDate>Thu, 29 Mar 2018 15:09:10 +0800</pubDate><description><![CDATA[Q:



A number sequence is defined as follows:

f(1) = 1, f(2) = 1, f(n) = (A * f(n - 1) + B * f(n - 2)) mod 7.

Given A, B, and n, you are to calculate the value of f(n).

Input:



The input consist...]]></description><category></category></item><item><title><![CDATA[保研经历回顾]]></title><link>https://blog.youkuaiyun.com/renxingkai/article/details/78218349</link><guid>https://blog.youkuaiyun.com/renxingkai/article/details/78218349</guid><author>renxingkai</author><pubDate>Thu, 12 Oct 2017 18:46:00 +0800</pubDate><description><![CDATA[保研经验回顾]]></description><category></category></item><item><title><![CDATA[LeetCode-121-Best Time to Buy and Sell Stock(最佳买卖股票的时间)]]></title><link>https://blog.youkuaiyun.com/renxingkai/article/details/77441852</link><guid>https://blog.youkuaiyun.com/renxingkai/article/details/77441852</guid><author>renxingkai</author><pubDate>Mon, 21 Aug 2017 09:21:38 +0800</pubDate><description><![CDATA[Q: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), de]]></description><category></category></item><item><title><![CDATA[《Python机器学习》笔记--感知机分类鸢尾花数据集]]></title><link>https://blog.youkuaiyun.com/renxingkai/article/details/77434749</link><guid>https://blog.youkuaiyun.com/renxingkai/article/details/77434749</guid><author>renxingkai</author><pubDate>Sun, 20 Aug 2017 21:41:41 +0800</pubDate><description><![CDATA[书籍参考机械工业出版社的《Python机器学习》，如下图: 
　　　　　　　　　　一、 
　　感知机模型是很基础的二类分类（判别）模型，主要采用了分离超平面的概念，其学习策略是极小化误分点到超平面距离，使用的学习算法为随机梯度下降算法。 
　　详细算法不再阐述，各位聚聚博客or书籍有大量讲解内容。 
二、 
　　此处使用sklearn包中的Perceptron（感知机）类来对sklearn包中提供的]]></description><category></category></item><item><title><![CDATA[LeetCode-118-Pascal's Triangle(帕斯卡的三角形)]]></title><link>https://blog.youkuaiyun.com/renxingkai/article/details/77399035</link><guid>https://blog.youkuaiyun.com/renxingkai/article/details/77399035</guid><author>renxingkai</author><pubDate>Sat, 19 Aug 2017 09:53:30 +0800</pubDate><description><![CDATA[Q:Given numRows, generate the first numRows of Pascal’s triangle.For example, given numRows = 5, 
Return[
     [1],
    [1,1],
   [1,2,1],
  [1,3,3,1],
 [1,4,6,4,1]
]Analysis:基础的算法问题，中间元素值等于“肩上”两元素之和。C]]></description><category></category></item><item><title><![CDATA[LeetCode-112-Path Sum(路径和)]]></title><link>https://blog.youkuaiyun.com/renxingkai/article/details/77285560</link><guid>https://blog.youkuaiyun.com/renxingkai/article/details/77285560</guid><author>renxingkai</author><pubDate>Thu, 17 Aug 2017 09:13:06 +0800</pubDate><description><![CDATA[Q:Given a binary tree and a sum, determine if the tree has a root-to-leaf path such that adding up all the values along the path equals the given sum.For example: 
Given the below binary tree and sum =]]></description><category></category></item><item><title><![CDATA[LeetCode-111-Minimum Depth of Binary Tree(二叉树的最短路径)]]></title><link>https://blog.youkuaiyun.com/renxingkai/article/details/77265684</link><guid>https://blog.youkuaiyun.com/renxingkai/article/details/77265684</guid><author>renxingkai</author><pubDate>Wed, 16 Aug 2017 19:07:32 +0800</pubDate><description><![CDATA[Q:Given a binary tree, find its minimum depth.The minimum depth is the number of nodes along the shortest path from the root node down to the nearest leaf node.Analysis:此题要求返回二叉树根结点到叶子结点最短的距离。可以采用递归的方法]]></description><category></category></item><item><title><![CDATA[LeetCode-110-Balanced Binary Tree(平衡二叉树)]]></title><link>https://blog.youkuaiyun.com/renxingkai/article/details/77119531</link><guid>https://blog.youkuaiyun.com/renxingkai/article/details/77119531</guid><author>renxingkai</author><pubDate>Sat, 12 Aug 2017 15:56:17 +0800</pubDate><description><![CDATA[Q:Given a binary tree, determine if it is height-balanced.For this problem, a height-balanced binary tree is defined as a binary tree in which the depth of the two subtrees of every node never differ b]]></description><category></category></item><item><title><![CDATA[LeetCode-108-Convert Sorted Array to Binary Search Tree(转化已排序数组到二叉排序树)]]></title><link>https://blog.youkuaiyun.com/renxingkai/article/details/77089779</link><guid>https://blog.youkuaiyun.com/renxingkai/article/details/77089779</guid><author>renxingkai</author><pubDate>Fri, 11 Aug 2017 10:24:51 +0800</pubDate><description><![CDATA[Q:Given an array where elements are sorted in ascending order, convert it to a height balanced BST.Analysis:题目要求将已升序的数组转化为二叉排序树；可以使用递归，二分的方法，进行生成二叉排序树。Code:/**
 * Definition for a binary tree node.
 *]]></description><category></category></item></channel></rss>