爱萌 发表于 2009-7-15 11:27 
ruiqwy 发表于 2009-7-15 11:20 
LZ的对各个软件的评价,某些是有道理的。但是这些比喻极其不恰当,反映出其对这些软件缺乏深刻的理解,也反映出其对统计学的认识缺乏深入的理解,LZ列出的很多统计软件,有些根本无法比较,做的方向不一,风马牛不相及。
不同的软件各有特长,根据自己的需要选择相应的软件,以及不同人有不同的使用习惯。一个人使用什么软件,从某种程度上反映出统计学的水平,一般来说,EXCEL等
呵呵,只能说明你自己的SAS水平不到家吗,我们的斑竹nkwilling已经给出了自己编写的朴素贝叶斯,我自己也成功的写了SVM,和其他我自己研究的方法
Apologize for not typing Chinese.
In my humble (or not so humble) opinion, ruiqwy was being too polite in commenting your post. I seriously doubt you ever finished any reasonable sized project with R/Splus or Matlab. This statistical packages comparison topic has been beaten to death. You can easily google or wiki key word like "comparison of statistical packages" to find more objective and thorough reviews. There is no need to reveal your lack of experience by thrashing certain packages.
What is the point of mentioning your being able to implement svm in sas? It is not about can or can not, it is about what is the right tool to finish the task at hand. Given enough time, I can implement svm in assembly code. According to your logic, that would make sas useless, right? :)