摘要: 原文地址:http://blog.csdn.net/jlei_apple/article/details/8168856这两天在看关于boosting算法时,看到一篇不错的文章讲bootstrap, jackknife, bagging, boosting, random forest 都有介绍,以... 阅读全文
posted @ 2014-07-07 11:25 快乐的Developer 阅读(205) 评论(0) 推荐(0) 编辑
摘要: 原文地址:http://hi.baidu.com/hehehehello/item/96cc42e45c16e7265a2d64eeGBDT(GradientBoostingDecisionTree)又 叫MART(MultipleAdditiveRegressionTree),是一种迭代的决策树算... 阅读全文
posted @ 2014-07-03 11:14 快乐的Developer 阅读(216) 评论(0) 推荐(0) 编辑
摘要: 原文地址:http://hi.baidu.com/yangchenhao/item/4d8d0b4a43067c0c6cc2f0ab如果你读过贝叶斯学习方面的书或者论文,想必是知道共轭先验这个名词的。贝叶斯学派和频率学派的区别之一是特别重视先验信息对于inference的影响,而引入先验信息的手段有“贝叶斯原则“(即把先验信息当着均匀分布)等四大类其中有重要影响的一类是:共轭先验现在假设我们有这样几类概率:P(\theta)(先验分布), p(\theta|X)(后验分布), p(X), p(X|\theta) (似然函数)它们之间的关系可以通过贝叶斯公式进行连接: 后验分布 = 似然函数* 阅读全文
posted @ 2013-09-24 14:41 快乐的Developer 阅读(372) 评论(0) 推荐(0) 编辑
摘要: Latent Semantic Analysis (LSA), also known as Latent Semantic Indexing (LSI) literally means analyzing documents to find the underlying meaning or concepts of those documents. If each word only meant one concept, and each concept was only described by one word, then LSA would be easy since there is 阅读全文
posted @ 2013-09-02 14:14 快乐的Developer 阅读(419) 评论(0) 推荐(0) 编辑
摘要: 引用:http://vaideeswaranr.blogspot.com/2013/02/using-microsoft-fakes-to-unit-test.htmlSunday, February 24, 2013Using Microsoft Fakes to Unit Test Entity FrameworkTechnorati Tags: Microsoft Fakes,Entity Framework,Unit TestingIf you have used Entity Framework, then I am sure you have read in various blo 阅读全文
posted @ 2013-04-02 23:47 快乐的Developer 阅读(423) 评论(0) 推荐(0) 编辑
摘要: more useful information extracted from the cited booknumeric(precision, scale) = decimal(precision, scale) - Exact Numeric Data Typesreal = float(24) - Approximate Numeric Data TypesAll exact numeric types always produce the same result, regardless of which kind of processor architecture is being us 阅读全文
posted @ 2013-01-06 14:22 快乐的Developer 阅读(302) 评论(0) 推荐(0) 编辑
摘要: R语言入门基础教程:常用运算函数。对一般数据进行运算的常用函数:1、round() #四舍五入例:x <- c(3.1416, 15.377, 269.7)round(x, 0) #保留整数位round(x, 2) #保留两位小数round(x, -1) #保留到十位2、signif() #取有效数字(跟学过的有效数字不是一个意思)例:略3、trunc() #取整 floor() #向下取整 ceiling() #向上取整例:xx <- c(3.60, 12.47, -3.60, -12.47)trunc(xx)floor(xx)ceiling(xx)4、logb(a, b) #以 阅读全文
posted @ 2012-12-18 22:51 快乐的Developer 阅读(377) 评论(0) 推荐(0) 编辑
摘要: 原理及实例: http://zh.wikipedia.org/wiki/%E8%B4%9D%E5%8F%B6%E6%96%AF%E5%AE%9A%E7%90%86 阅读全文
posted @ 2012-11-06 14:56 快乐的Developer 阅读(166) 评论(0) 推荐(0) 编辑
摘要: http://files.cnblogs.com/liutingrex/%E4%B8%BB%E5%85%83%E5%88%86%E6%9E%90PCA%E7%90%86%E8%AE%BA%E5%88%86%E6%9E%90%E5%8F%8A%E5%BA%94%E7%94%A8.zip 阅读全文
posted @ 2012-11-01 17:35 快乐的Developer 阅读(269) 评论(0) 推荐(0) 编辑
摘要: 数据预处理主要步骤: 聚集;抽样;维规约;特征子集选择;特征创建;离散化和二元化;变量变换。 阅读全文
posted @ 2012-10-30 16:39 快乐的Developer 阅读(165) 评论(0) 推荐(0) 编辑