注册 登录  
 加关注
   显示下一条  |  关闭
温馨提示!由于新浪微博认证机制调整,您的新浪微博帐号绑定已过期,请重新绑定!立即重新绑定新浪微博》  |  关闭

数据挖掘

学习数据挖掘

 
 
 

日志

 
 

罗辑回归  

2014-04-09 16:16:19|  分类: SAS |  标签: |举报 |字号 订阅

  下载LOFTER 我的照片书  |

1. Introduction to Logistic Regression Model
English Materials:
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065119/ (Theory)
http://www.ats.ucla.edu/stat/sas/topics/logistic.htm (Practice using SAS)
https://support.sas.com/edu/schedules.html?ctry=us&id=52 (Training provided by SAS)
http://www.docin.com/p-81251422.html (SAS Course Notes of Predicitve Modeling Using Logistic Regression)
http://v.163.com/special/sp/statisticalanalysisofcategoricaldata.html (Public Class by Nicholas P. JEWELL, University of California, Berkeley)
 
Chinese Materials: 
http://www.cnblogs.com/biyeymyhjob/archive/2012/07/18/2595410.html
http://blog.csdn.net/abcjennifer/article/details/7716281#comments
http://www.douban.com/note/230863157/
  
2. How to build model using SAS?
http://www.ats.ucla.edu/stat/sas/dae/logit.htm 
 
3. Data Process
However, as most people who have developed predictive regression or other behavioral models are aware, the bulk of the time spent in developing models isn’t in the final production of the regression coefficients, but in variable selection and preparing the input variables (e.g. clustering levels, imputing missing values, etc.).
 
Imputation
http://www.songhuiming.com/2012/09/study-notes1-predictive-modeling-with.html
 
Reduce variables (Principle component analysis for numeric variables)
http://www.nesug.org/Proceedings/nesug12/sa/sa09.pdf
http://support.sas.com/resources/papers/sgf2008/2stagecluster.pdf
 
Reduce the number of levels in the categorical variables (Greenacre's method)
http://www.songhuiming.com/2012/09/study-notes2-predictive-modeling-with.html
http://www2.sas.com/proceedings/sugi31/079-31.pdf
http://www.mwsug.org/proceedings/2012/SA/MWSUG-2012-SA03.pdf
 
Complete separation (quasi-complete separation)
http://www.ats.ucla.edu/stat/mult_pkg/faq/general/complete_separation_logit_models.htm
 
4. Score
http://blog.sina.com.cn/s/blog_8db50cf70101bg80.html
  评论这张
 
阅读(208)| 评论(0)
推荐 转载

历史上的今天

评论

<#--最新日志,群博日志--> <#--推荐日志--> <#--引用记录--> <#--博主推荐--> <#--随机阅读--> <#--首页推荐--> <#--历史上的今天--> <#--被推荐日志--> <#--上一篇,下一篇--> <#-- 热度 --> <#-- 网易新闻广告 --> <#--右边模块结构--> <#--评论模块结构--> <#--引用模块结构--> <#--博主发起的投票-->
 
 
 
 
 
 
 
 
 
 
 
 
 
 

页脚

网易公司版权所有 ©1997-2017