wssjdysf1的专栏

TextRank代码实现package programimport algorithm.TextRankimport org.apache.spark.sql.SaveModeimport util.{SegmentWordUtil, SparkSessionBase}import scala.collection.mutableimport scala.collection.mutable.ListBufferobject ComputeTextR...

https://github.com/bjmashibing/Recommender/blob/master/%E6%8E%A8%E8%8D%90%E7%B3%BB%E7%BB%9F05-%E5%9F%BA%E4%BA%8E%E8%8A%82%E7%9B%AE%E7%9A%84%E6%8E%A8%E8%8D%90%E7%B3%BB%E7%BB%9F%E6%9E%B6%E6%9E%84%E5%89%96%E6%9E%90/%E5%9F%BA%E4%BA%8E%E8%8A%82%E7%9B%AE%E7%9A%8

https://download.youkuaiyun.com/download/wssjdysf1/14920495https://github.com/bjmashibing/Recommender/blob/master/%E6%8E%A8%E8%8D%90%E7%B3%BB%E7%BB%9F04-%E5%AE%9E%E7%8E%B0%E5%9C%A8%E7%BA%BF%E6%8E%A8%E8%8D%90%E5%BE%AE%E6%9C%8D%E5%8A%A1/Dubbo.docx...

https://raw.githubusercontent.com/bjmashibing/Recommender/master/%E6%8E%A8%E8%8D%90%E7%B3%BB%E7%BB%9F02-%E6%9E%84%E5%BB%BA%E6%8E%92%E5%BA%8F%E6%A8%A1%E5%9E%8B%E8%AE%AD%E7%BB%83%E9%9B%86/%E6%8E%A8%E8%8D%90%E7%B3%BB%E7%BB%9F-%E6%9E%84%E5%BB%BA%E8%AE%AD...

https://github.com/bjmashibing

https://raw.githubusercontent.com/bjmashibing/Machine-Learning/master/%E6%9C%BA%E5%99%A8%E5%AD%A6%E4%B9%A013-%E5%86%B3%E7%AD%96%E6%A0%91%E4%B8%8E%E9%9A%8F%E6%9C%BA%E6%A3%AE%E6%9E%97/%E5%86%B3%E7%AD%96%E6%A0%91%E4%B8%8E%E9%9A%8F%E6%9C%BA%E6%A3%AE%E6%9E%97.p

https://github.com/bjmashibing/Machine-Learning/blob/master/%E6%9C%BA%E5%99%A8%E5%AD%A6%E4%B9%A004-KNN%E7%AE%97%E6%B3%95/KNN%E7%AE%97%E6%B3%95.png

https://raw.githubusercontent.com/bjmashibing/Machine-Learning/master/%E6%9C%BA%E5%99%A8%E5%AD%A6%E4%B9%A010-%E9%80%BB%E8%BE%91%E5%9B%9E%E5%BD%92%E7%AE%97%E6%B3%95%E8%B0%83%E4%BC%98%E5%8F%8A%E5%9C%BA%E6%99%AF%E5%88%86%E6%9E%90/%E9%80%BB%E8%BE%91%E5%9B%9E%E

讲课PPT：https://download.youkuaiyun.com/download/wssjdysf1/14369514讲课笔记：

决策树的优缺点优点：计算复杂度不高，输出结果易于理解，对中间值的缺失不敏感，可以处理不相关特征数据。缺点：可能会产生过度匹配问题。适用数据类型：数值型和标称型。决策树的大致流程1.根据信息增益(熵)，寻找划分数据集的最好特征。2.划分数据集3.创建分支节点(建模)基础概念https://shuwoom.com/?p=1452实战项目https://www...

优缺点优点：缺点：适用数据类型：大致流程基础概念实战项目Spark实现

优缺点优点：在数据较少的情况下仍然有效，可以处理多类别的问题。缺点：对于输入数据的准备方式较为敏感。适用数据类型：标称型数据。大致流程基础概念https://www.cnblogs.com/sxron/p/5452821.html【简单看看】https://blog.youkuaiyun.com/AMDS123/article/details/70173402#reply%2...

阿里巴巴企业文化https://zhidao.baidu.com/question/537553305.htmlOpenJDK源码【看本地方法】http://hg.openjdk.java.net/jdk8u/jdk8u/jdk/file/JDK官方文档https://docs.oracle.com/javase/specs/jvms/se8/html/index.htm...

knn的优缺点优点：精度高，对异常值不敏感，无数据输入假定。缺点：计算复杂度高，空间复杂度高。适用数据范围：数值型和标称型。knn的大致流程1.计算未知点与数据集中所有点的距离。2.按距离从小到大排序。【距离值越小，越相似】3.取前k个距离相似的点。4.统计每个点的分类标签。【判断分类标签，哪个标签出现的次数多，未知点就属于哪种分类】基础概念https://...